5.0.1: Why can't I specify more than 2896 centroids when creating a new data bank? See the answer to this question in the archived FAQs. 5.0.2: Do my centroid numbers need to be between 1 and the maximum number of zones permitted by my licence? No. The node numbers, including the centroids, need not be consecutive and may have up to 6 digits. However, if your node numbers have more than 4 digits, be sure to set the appropriate number of digits in Module 1.23, using option 7 of the Primary select. 5.0.3: I modified the device table file but the modification does not seem to have been included in the data bank. This is because the new file was not read into the data bank, either because switch 10 was not set, or because the wrong file was read. In order to input the contents of the new device table file, perform the following: Enter EMME/2 At the main menu level, set switch 10 ON Make sure that the file to be read is correctly defined using the files command at the main menu level. If the output of this command is: ... batchin: ... The new device table should be named d002.in and should be in the current application directory. If this is NOT the case, set the file name using the menu command batchin=? and specify the correct file name, including the complete path if the file is not in the current application directory. Exit EMME/2 (using q or 9) Enter EMME/2 using the command emme2 (NOT followed by a module number). 5.0.4: Is it possible to use coordinates with 7 significant digits? It is possible but the window coordinates shown on network plots may be displayed as *******. Thirteen characters are allocated to display the window coordinates, with a maximum of 7 characters for each number, including any necessary decimal points and minus signs. A coordinate with decimals will be rounded to fit the 7 character field width, if possible. Any coordinate which cannot be displayed using 7 characters, for example a positive integer with more than 7 digits, will always be displayed as *******. 5.0.5: Before leaving a module I tried to look at the batchout file using for example, "~! type batchout", but the file was empty. Why? The batchout file is closed only when leaving the module. Before leaving the module, the contents of the batchout file may be incomplete because the file buffer is only emptied when the file is closed. Note that the same is true for the reports, plots and errors files. Therefore, always quit a module before trying to use operating system commands for file access via the command escape (~ !). 5.1 Network Modifications 5.1.1: Why do I get errors when trying to read a d211.in file containing node numbers with more than 4 digits? By default, the number of digits for node numbers is limited to 4. However, it is possible to use node numbers of up to 6 digits. To change the default, use option 7 of Module 1.23. (See also Question 5.0.2). 5.1.2: The length of my links was entered using *. Will the length be updated automatically if I modify the coordinates of a node? No. You will have to modify the attribute len of the links using that node. This can be done in Module 2.12, in Module 2.11 or in Module 2.41, where the Euclidean distance between the I- and J-node can be computed using the expression: sqrt((xi-xj)^2+(yi-yj)^2)* scale factor (the scale factor is the length of a coordinate unit as specified in Module 1.23). The links can be selected using: i=node number or j=node number 5.1.3: How can I modify a link attribute? There are many possibilities. If you have only a few links to modify, you can use the graphic worksheet (Module 2.12). You can punch the links to be modified (Module 2.14), edit the batchout file and read the new file in batch (Module 2.11). If the new values are contained in a file which is NOT in d211.in format, use option 3 of Module 2.41 to input the values in batch. If the modifications are systematic, you can use Module 2.41 and save the result of the expression (which can be a constant) in an attribute. In the link selection, be sure to include only the links to be modified. 5.1.4: I would like to initialize my base network and read a new d211.in file but I am not allowed to because there are turns (and/or transit lines) already defined. How can I read my file without losing my turns (and/or transit lines)? This is not possible due to the hierarchy of the data bank: modes (and vehicles), nodes, links, turns, transit lines (see the EMME/2 User's Manual, Chapter 3, Figure 3-1). You should thus proceed as follows. Punch all turns using Module 2.31, then delete all turns. Punch all transit lines using Module 2.24, delete all transit lines. Initialize the network, read the new d211.in file and read back the turn and transit line batchout files. Turns and transit lines which are not compatible with the new network will generate errors. 5.2 Network Plots 5.2.1: How do I change the size of arrowheads on network plots? The size of the arrowheads can be modified in Module 1.23. The default size is 2.0. Using 0 will give links with no arrowheads. 5.2.2: How can I obtain a plot where nodes are connected by only one vector (even if two-way links exist) and with no arrowhead? Use Module 1.23 to set the size of the arrowheads to 0. In the desired modules, set the module parameter Offset to the right for links to 0. 5.2.3: Once I have a plot, how can I zoom out? Use the graphic command w followed by a number between 2 and 9. This will result in a window centered on the same node but covering an area 2 to 9 times larger. 5.2.4: In Module 2.13, how can I display the value of link or node attribute with a fixed number of decimals? You can create a new extra attribute and compute the value with the required number of decimals in Module 2.41. For example, to display the value of ul3 with 2 decimals, perform the following calculation: @att=nint(ul3*100)/100 and then, display the value of @att in Module 2.13. Starting with Release 9, you can also use the new sub-option 3 (several user defined link attributes) of option 6 (user defined data) to customize the display of link attributes, including the specification of the number of decimals. When selecting this option, you can fix the number of digits after the decimal point by specifying the selected number (y) as the second value for the Enter: Color index for layer N question following the specification of the attribute (see the EMME/2 User's Manual, Module 2.13, option 6). 5.2.5: Is it possible to display node AND link data on the same plot? No. However, it is possible to superimpose 2 plots by doing the following: set switch 18 ON to suppress the standard plot titles generate a plot with the node data outside EMME/2, transform the plot into an annotation (annotx) using GPL and the ANNOT.GPL format file (gpl -f annot.gpl -o annotx plots) generate a plot with the link data and recall annotation x (using the previous window). 5.2.6: Why is it that, when I use link color index in Module 6.12, I do not see all my links? Check that the "missing" links have a positive color index. Except in graphic worksheets, all links with a color index less than or equal to 0 are not drawn or drawn in the background color. 5.2.7: Can the information added to the plot with interactive commands in Modules 2.15, 2.43 and 6.15 be written to the plot file? Normally, this information is not sent to the plot file. However, if switch 14 is ON, all graphic output which is transmitted to the terminal is also copied to the plot file, using the terminal device definition. Note that this mode of operation is not standard and is NOT recommended for general use since it creates plot files that may be difficult to process with GPL, GPR or E2PLOTS. You can also use the Print Page option in EMTOOLW to print all the information actually on the screen. 5.2.8: How can I obtain a plot where node numbers (labels) are displayed for centroids only? Starting with Release 9, you can use the new sub-option 3 (several user defined link attributes) of option 6 (user defined data) to customize the display of link attributes, together with the specification of the subset of nodes for which the node number or label is displayed, by choosing option 3 of the following new Select question: Select: Type of node number display 1= display no node numbers 2= display all node numbers 3= display selected node numbers For more details, refer to the EMME/2 User's Manual, Module 2.13, option 6. For releases prior to Release 9, the cent2ann.awk script, available in the EMME/2, STAN Scripts section, can be used to generate an annotation file containing only the centroid numbers or labels, using the network input file (d211.in) as input. 5.3 Network and Matrix Calculators 5.3.1: How can I set color indices for different types of links? The best way is to use the network calculator (Module 2.41) to set the values in a link attribute. For example, if you want to color the connectors in green, the expression would simply be 3 (for color index 3) and, to apply the result to the connectors only, the link selection would be: ci=1 |cj=1 Note that the link attribute must first be initialized to the desired (positive) color for all other links since links with a color index of 0 are not displayed. 5.3.2: Can temporary link attributes tmplx be used in selections? No. However, it is possible to include conditions within the expression in order to "simulate" the selection. For example, if you want to change the length of the links only for those links having tmpl1 greater than 0, the expression would be: len * (tmpl1 <= 0) + expression * (tmpl1 > 0) It is also possible to replace temporary attributes by extra attributes, which can be used in the selections. 5.3.3: How can I write an expression that ensures a negative value is replaced by 0? Use the .max. operator: (expression) .max.0. 5.3.4: In the network and matrix calculators, why are the results saved when I send the report to the printer but not when I send the report to the terminal? Probably because you stopped the display of the report prematurely with a r or q. If you do not want to look at the remainder of the report, type l to skip to the end of the report. 5.4 Matrices 5.4.1: Why do some of my matrices have a "/c" flag? Probably because a transit assignment did not end normally. During a transit assignment, matrices are transformed to columnwise format in order to work destination by destination. At the end of the assignment, the matrices are retransformed to rowwise format. If the transit assignment does not end normally, the matrices will stay columnwise. Since working with columnwise matrices will increase the matrix access time drastically, you should change the storage type of columnwise matrices by using Module 3.12, option 3 (modify a matrix). 5.4.2: Matrix calculations are very slow. Is there a way to speed this up? You could do the following: Check that the matrices used in the calculations are not columnwise (see also Question 5.4.1). When running a macro with many calculations, suppress programmed pauses (~o|32). On Intel platforms, try to perform disk defragmentation. If you are using DOS or Windows 3.x, verify the size of your disk cache (Smartdrive). On UNIX platforms, try to use the data bank locally and not via a network. 5.4.3: How can I list several origin and destination matrices, excluding the zones for which the value is 0 for all matrices? Prepare an origin matrix that will be used as a constraint matrix. This matrix should contain the sum of all origin and destination matrices (destination matrices should be transposed!). For example: mo99 = mo1 + mo2 + md3'. If matrices contain negative values, the expression should be: abs(mo1) + abs(mo2) + abs(md3'). 5.5 Functions 5.5.1: How can I try different values for a constant that appears in my volume-delay function without entering the complete function each time? Put the value in a scalar matrix n and then refer to it in the function using ms(n). 5.5.2: One of the sample data files provided with the EMME/2 system contains functions. Are they general enough to be used with any data bank? No. These functions are specific to the Winnipeg data and are included solely for demonstration purposes. 5.6 Assignments 5.6.1: After an auto assignment, are the times on the connectors included in the auto time saved as a matrix? Yes. The time will depend on the functions for the connectors. 5.6.2: Is it possible to stop an auto assignment before one of the specified stopping criteria is met? Yes. Use Ctrl-C. The assignment will stop at the end of the next iteration. 5.6.3: How can I identify the non-assigned trips after an assignment? You can output the demand matrix using the time matrix as a constraint matrix with the constraint interval: (0.001,9999,exclude). This will include the trips on the diagonal (travel time of zero) and the non-connected trips (travel time equal to infinity). 5.6.4: I have a travel time matrix. How can I obtain the minimum travel time for each origin? In Module 3.21, perform a matrix computation and save the result in an origin matrix. Use the travel time matrix for the expression and as a constraint matrix (to exclude 0 values). Use the .min. operator for aggregation over destinations. 5.6.5: How can I perform an additional options auto assignment AND consider transit vehicles as additional volumes at the same time? You can compute the auto equivalent of the transit vehicles in Module 2.41 (vauteq*60/hdwy) and store the results in link and turn user data. Then, replace volad by ulN in the volume-delay functions, and pvolad by upN in the penalty functions. 5.6.6: How can I optimize the speed of my assignment? You can suppress the computation of the objective function by setting the relative gap to a negative value (for example, -1). You should also simplify your function expressions as much as possible. For example, replace sub-expressions such as constantA/constantB by constantC and use put() and get() functions whenever possible. 5.7 Zone Systems 5.7.1: Can I use different zone systems in the same data bank? It is possible but not recommended. It is better to have different data banks for different zone systems. Refer to the EMME/2 User's Manual, Chapter 3, Section 3.2.2 Matrices and scenarios. 5.7.2: I need more detailed zones in some parts of my network and more aggregate data in other parts. What is the best way to modify my zone system? Refer to the EMME/2 News of March 87 - Moving matrix data between different zone systems. The basic idea is to: create 2 data banks, 1 for each zone system in each data bank, create an ensemble that will be used to define the correspondence between the 2 zone systems punch and read the matrices using this ensemble (further manipulations might be required) 5.7.3: How can I increase the number of centroids in my data bank? While it is possible to increase most of the data bank dimensions by using the macro modifydb, it is not possible to increase the number of centroids. If this is necessary, the only solution is to create a new data bank with the new dimensions. Data from the previous data bank can then be imported using Module 1.31.
See the answer to this question in the archived FAQs.
5.0.2: Do my centroid numbers need to be between 1 and the maximum number of zones permitted by my licence? No. The node numbers, including the centroids, need not be consecutive and may have up to 6 digits. However, if your node numbers have more than 4 digits, be sure to set the appropriate number of digits in Module 1.23, using option 7 of the Primary select. 5.0.3: I modified the device table file but the modification does not seem to have been included in the data bank. This is because the new file was not read into the data bank, either because switch 10 was not set, or because the wrong file was read. In order to input the contents of the new device table file, perform the following: Enter EMME/2 At the main menu level, set switch 10 ON Make sure that the file to be read is correctly defined using the files command at the main menu level. If the output of this command is: ... batchin: ... The new device table should be named d002.in and should be in the current application directory. If this is NOT the case, set the file name using the menu command batchin=? and specify the correct file name, including the complete path if the file is not in the current application directory. Exit EMME/2 (using q or 9) Enter EMME/2 using the command emme2 (NOT followed by a module number). 5.0.4: Is it possible to use coordinates with 7 significant digits? It is possible but the window coordinates shown on network plots may be displayed as *******. Thirteen characters are allocated to display the window coordinates, with a maximum of 7 characters for each number, including any necessary decimal points and minus signs. A coordinate with decimals will be rounded to fit the 7 character field width, if possible. Any coordinate which cannot be displayed using 7 characters, for example a positive integer with more than 7 digits, will always be displayed as *******. 5.0.5: Before leaving a module I tried to look at the batchout file using for example, "~! type batchout", but the file was empty. Why? The batchout file is closed only when leaving the module. Before leaving the module, the contents of the batchout file may be incomplete because the file buffer is only emptied when the file is closed. Note that the same is true for the reports, plots and errors files. Therefore, always quit a module before trying to use operating system commands for file access via the command escape (~ !). 5.1 Network Modifications 5.1.1: Why do I get errors when trying to read a d211.in file containing node numbers with more than 4 digits? By default, the number of digits for node numbers is limited to 4. However, it is possible to use node numbers of up to 6 digits. To change the default, use option 7 of Module 1.23. (See also Question 5.0.2). 5.1.2: The length of my links was entered using *. Will the length be updated automatically if I modify the coordinates of a node? No. You will have to modify the attribute len of the links using that node. This can be done in Module 2.12, in Module 2.11 or in Module 2.41, where the Euclidean distance between the I- and J-node can be computed using the expression: sqrt((xi-xj)^2+(yi-yj)^2)* scale factor (the scale factor is the length of a coordinate unit as specified in Module 1.23). The links can be selected using: i=node number or j=node number 5.1.3: How can I modify a link attribute? There are many possibilities. If you have only a few links to modify, you can use the graphic worksheet (Module 2.12). You can punch the links to be modified (Module 2.14), edit the batchout file and read the new file in batch (Module 2.11). If the new values are contained in a file which is NOT in d211.in format, use option 3 of Module 2.41 to input the values in batch. If the modifications are systematic, you can use Module 2.41 and save the result of the expression (which can be a constant) in an attribute. In the link selection, be sure to include only the links to be modified. 5.1.4: I would like to initialize my base network and read a new d211.in file but I am not allowed to because there are turns (and/or transit lines) already defined. How can I read my file without losing my turns (and/or transit lines)? This is not possible due to the hierarchy of the data bank: modes (and vehicles), nodes, links, turns, transit lines (see the EMME/2 User's Manual, Chapter 3, Figure 3-1). You should thus proceed as follows. Punch all turns using Module 2.31, then delete all turns. Punch all transit lines using Module 2.24, delete all transit lines. Initialize the network, read the new d211.in file and read back the turn and transit line batchout files. Turns and transit lines which are not compatible with the new network will generate errors. 5.2 Network Plots 5.2.1: How do I change the size of arrowheads on network plots? The size of the arrowheads can be modified in Module 1.23. The default size is 2.0. Using 0 will give links with no arrowheads. 5.2.2: How can I obtain a plot where nodes are connected by only one vector (even if two-way links exist) and with no arrowhead? Use Module 1.23 to set the size of the arrowheads to 0. In the desired modules, set the module parameter Offset to the right for links to 0. 5.2.3: Once I have a plot, how can I zoom out? Use the graphic command w followed by a number between 2 and 9. This will result in a window centered on the same node but covering an area 2 to 9 times larger. 5.2.4: In Module 2.13, how can I display the value of link or node attribute with a fixed number of decimals? You can create a new extra attribute and compute the value with the required number of decimals in Module 2.41. For example, to display the value of ul3 with 2 decimals, perform the following calculation: @att=nint(ul3*100)/100 and then, display the value of @att in Module 2.13. Starting with Release 9, you can also use the new sub-option 3 (several user defined link attributes) of option 6 (user defined data) to customize the display of link attributes, including the specification of the number of decimals. When selecting this option, you can fix the number of digits after the decimal point by specifying the selected number (y) as the second value for the Enter: Color index for layer N question following the specification of the attribute (see the EMME/2 User's Manual, Module 2.13, option 6). 5.2.5: Is it possible to display node AND link data on the same plot? No. However, it is possible to superimpose 2 plots by doing the following: set switch 18 ON to suppress the standard plot titles generate a plot with the node data outside EMME/2, transform the plot into an annotation (annotx) using GPL and the ANNOT.GPL format file (gpl -f annot.gpl -o annotx plots) generate a plot with the link data and recall annotation x (using the previous window). 5.2.6: Why is it that, when I use link color index in Module 6.12, I do not see all my links? Check that the "missing" links have a positive color index. Except in graphic worksheets, all links with a color index less than or equal to 0 are not drawn or drawn in the background color. 5.2.7: Can the information added to the plot with interactive commands in Modules 2.15, 2.43 and 6.15 be written to the plot file? Normally, this information is not sent to the plot file. However, if switch 14 is ON, all graphic output which is transmitted to the terminal is also copied to the plot file, using the terminal device definition. Note that this mode of operation is not standard and is NOT recommended for general use since it creates plot files that may be difficult to process with GPL, GPR or E2PLOTS. You can also use the Print Page option in EMTOOLW to print all the information actually on the screen. 5.2.8: How can I obtain a plot where node numbers (labels) are displayed for centroids only? Starting with Release 9, you can use the new sub-option 3 (several user defined link attributes) of option 6 (user defined data) to customize the display of link attributes, together with the specification of the subset of nodes for which the node number or label is displayed, by choosing option 3 of the following new Select question: Select: Type of node number display 1= display no node numbers 2= display all node numbers 3= display selected node numbers For more details, refer to the EMME/2 User's Manual, Module 2.13, option 6. For releases prior to Release 9, the cent2ann.awk script, available in the EMME/2, STAN Scripts section, can be used to generate an annotation file containing only the centroid numbers or labels, using the network input file (d211.in) as input. 5.3 Network and Matrix Calculators 5.3.1: How can I set color indices for different types of links? The best way is to use the network calculator (Module 2.41) to set the values in a link attribute. For example, if you want to color the connectors in green, the expression would simply be 3 (for color index 3) and, to apply the result to the connectors only, the link selection would be: ci=1 |cj=1 Note that the link attribute must first be initialized to the desired (positive) color for all other links since links with a color index of 0 are not displayed. 5.3.2: Can temporary link attributes tmplx be used in selections? No. However, it is possible to include conditions within the expression in order to "simulate" the selection. For example, if you want to change the length of the links only for those links having tmpl1 greater than 0, the expression would be: len * (tmpl1 <= 0) + expression * (tmpl1 > 0) It is also possible to replace temporary attributes by extra attributes, which can be used in the selections. 5.3.3: How can I write an expression that ensures a negative value is replaced by 0? Use the .max. operator: (expression) .max.0. 5.3.4: In the network and matrix calculators, why are the results saved when I send the report to the printer but not when I send the report to the terminal? Probably because you stopped the display of the report prematurely with a r or q. If you do not want to look at the remainder of the report, type l to skip to the end of the report. 5.4 Matrices 5.4.1: Why do some of my matrices have a "/c" flag? Probably because a transit assignment did not end normally. During a transit assignment, matrices are transformed to columnwise format in order to work destination by destination. At the end of the assignment, the matrices are retransformed to rowwise format. If the transit assignment does not end normally, the matrices will stay columnwise. Since working with columnwise matrices will increase the matrix access time drastically, you should change the storage type of columnwise matrices by using Module 3.12, option 3 (modify a matrix). 5.4.2: Matrix calculations are very slow. Is there a way to speed this up? You could do the following: Check that the matrices used in the calculations are not columnwise (see also Question 5.4.1). When running a macro with many calculations, suppress programmed pauses (~o|32). On Intel platforms, try to perform disk defragmentation. If you are using DOS or Windows 3.x, verify the size of your disk cache (Smartdrive). On UNIX platforms, try to use the data bank locally and not via a network. 5.4.3: How can I list several origin and destination matrices, excluding the zones for which the value is 0 for all matrices? Prepare an origin matrix that will be used as a constraint matrix. This matrix should contain the sum of all origin and destination matrices (destination matrices should be transposed!). For example: mo99 = mo1 + mo2 + md3'. If matrices contain negative values, the expression should be: abs(mo1) + abs(mo2) + abs(md3'). 5.5 Functions 5.5.1: How can I try different values for a constant that appears in my volume-delay function without entering the complete function each time? Put the value in a scalar matrix n and then refer to it in the function using ms(n). 5.5.2: One of the sample data files provided with the EMME/2 system contains functions. Are they general enough to be used with any data bank? No. These functions are specific to the Winnipeg data and are included solely for demonstration purposes. 5.6 Assignments 5.6.1: After an auto assignment, are the times on the connectors included in the auto time saved as a matrix? Yes. The time will depend on the functions for the connectors. 5.6.2: Is it possible to stop an auto assignment before one of the specified stopping criteria is met? Yes. Use Ctrl-C. The assignment will stop at the end of the next iteration. 5.6.3: How can I identify the non-assigned trips after an assignment? You can output the demand matrix using the time matrix as a constraint matrix with the constraint interval: (0.001,9999,exclude). This will include the trips on the diagonal (travel time of zero) and the non-connected trips (travel time equal to infinity). 5.6.4: I have a travel time matrix. How can I obtain the minimum travel time for each origin? In Module 3.21, perform a matrix computation and save the result in an origin matrix. Use the travel time matrix for the expression and as a constraint matrix (to exclude 0 values). Use the .min. operator for aggregation over destinations. 5.6.5: How can I perform an additional options auto assignment AND consider transit vehicles as additional volumes at the same time? You can compute the auto equivalent of the transit vehicles in Module 2.41 (vauteq*60/hdwy) and store the results in link and turn user data. Then, replace volad by ulN in the volume-delay functions, and pvolad by upN in the penalty functions. 5.6.6: How can I optimize the speed of my assignment? You can suppress the computation of the objective function by setting the relative gap to a negative value (for example, -1). You should also simplify your function expressions as much as possible. For example, replace sub-expressions such as constantA/constantB by constantC and use put() and get() functions whenever possible. 5.7 Zone Systems 5.7.1: Can I use different zone systems in the same data bank? It is possible but not recommended. It is better to have different data banks for different zone systems. Refer to the EMME/2 User's Manual, Chapter 3, Section 3.2.2 Matrices and scenarios. 5.7.2: I need more detailed zones in some parts of my network and more aggregate data in other parts. What is the best way to modify my zone system? Refer to the EMME/2 News of March 87 - Moving matrix data between different zone systems. The basic idea is to: create 2 data banks, 1 for each zone system in each data bank, create an ensemble that will be used to define the correspondence between the 2 zone systems punch and read the matrices using this ensemble (further manipulations might be required) 5.7.3: How can I increase the number of centroids in my data bank? While it is possible to increase most of the data bank dimensions by using the macro modifydb, it is not possible to increase the number of centroids. If this is necessary, the only solution is to create a new data bank with the new dimensions. Data from the previous data bank can then be imported using Module 1.31.
No. The node numbers, including the centroids, need not be consecutive and may have up to 6 digits. However, if your node numbers have more than 4 digits, be sure to set the appropriate number of digits in Module 1.23, using option 7 of the Primary select.
5.0.3: I modified the device table file but the modification does not seem to have been included in the data bank. This is because the new file was not read into the data bank, either because switch 10 was not set, or because the wrong file was read. In order to input the contents of the new device table file, perform the following: Enter EMME/2 At the main menu level, set switch 10 ON Make sure that the file to be read is correctly defined using the files command at the main menu level. If the output of this command is: ... batchin: ... The new device table should be named d002.in and should be in the current application directory. If this is NOT the case, set the file name using the menu command batchin=? and specify the correct file name, including the complete path if the file is not in the current application directory. Exit EMME/2 (using q or 9) Enter EMME/2 using the command emme2 (NOT followed by a module number). 5.0.4: Is it possible to use coordinates with 7 significant digits? It is possible but the window coordinates shown on network plots may be displayed as *******. Thirteen characters are allocated to display the window coordinates, with a maximum of 7 characters for each number, including any necessary decimal points and minus signs. A coordinate with decimals will be rounded to fit the 7 character field width, if possible. Any coordinate which cannot be displayed using 7 characters, for example a positive integer with more than 7 digits, will always be displayed as *******. 5.0.5: Before leaving a module I tried to look at the batchout file using for example, "~! type batchout", but the file was empty. Why? The batchout file is closed only when leaving the module. Before leaving the module, the contents of the batchout file may be incomplete because the file buffer is only emptied when the file is closed. Note that the same is true for the reports, plots and errors files. Therefore, always quit a module before trying to use operating system commands for file access via the command escape (~ !). 5.1 Network Modifications 5.1.1: Why do I get errors when trying to read a d211.in file containing node numbers with more than 4 digits? By default, the number of digits for node numbers is limited to 4. However, it is possible to use node numbers of up to 6 digits. To change the default, use option 7 of Module 1.23. (See also Question 5.0.2). 5.1.2: The length of my links was entered using *. Will the length be updated automatically if I modify the coordinates of a node? No. You will have to modify the attribute len of the links using that node. This can be done in Module 2.12, in Module 2.11 or in Module 2.41, where the Euclidean distance between the I- and J-node can be computed using the expression: sqrt((xi-xj)^2+(yi-yj)^2)* scale factor (the scale factor is the length of a coordinate unit as specified in Module 1.23). The links can be selected using: i=node number or j=node number 5.1.3: How can I modify a link attribute? There are many possibilities. If you have only a few links to modify, you can use the graphic worksheet (Module 2.12). You can punch the links to be modified (Module 2.14), edit the batchout file and read the new file in batch (Module 2.11). If the new values are contained in a file which is NOT in d211.in format, use option 3 of Module 2.41 to input the values in batch. If the modifications are systematic, you can use Module 2.41 and save the result of the expression (which can be a constant) in an attribute. In the link selection, be sure to include only the links to be modified. 5.1.4: I would like to initialize my base network and read a new d211.in file but I am not allowed to because there are turns (and/or transit lines) already defined. How can I read my file without losing my turns (and/or transit lines)? This is not possible due to the hierarchy of the data bank: modes (and vehicles), nodes, links, turns, transit lines (see the EMME/2 User's Manual, Chapter 3, Figure 3-1). You should thus proceed as follows. Punch all turns using Module 2.31, then delete all turns. Punch all transit lines using Module 2.24, delete all transit lines. Initialize the network, read the new d211.in file and read back the turn and transit line batchout files. Turns and transit lines which are not compatible with the new network will generate errors. 5.2 Network Plots 5.2.1: How do I change the size of arrowheads on network plots? The size of the arrowheads can be modified in Module 1.23. The default size is 2.0. Using 0 will give links with no arrowheads. 5.2.2: How can I obtain a plot where nodes are connected by only one vector (even if two-way links exist) and with no arrowhead? Use Module 1.23 to set the size of the arrowheads to 0. In the desired modules, set the module parameter Offset to the right for links to 0. 5.2.3: Once I have a plot, how can I zoom out? Use the graphic command w followed by a number between 2 and 9. This will result in a window centered on the same node but covering an area 2 to 9 times larger. 5.2.4: In Module 2.13, how can I display the value of link or node attribute with a fixed number of decimals? You can create a new extra attribute and compute the value with the required number of decimals in Module 2.41. For example, to display the value of ul3 with 2 decimals, perform the following calculation: @att=nint(ul3*100)/100 and then, display the value of @att in Module 2.13. Starting with Release 9, you can also use the new sub-option 3 (several user defined link attributes) of option 6 (user defined data) to customize the display of link attributes, including the specification of the number of decimals. When selecting this option, you can fix the number of digits after the decimal point by specifying the selected number (y) as the second value for the Enter: Color index for layer N question following the specification of the attribute (see the EMME/2 User's Manual, Module 2.13, option 6). 5.2.5: Is it possible to display node AND link data on the same plot? No. However, it is possible to superimpose 2 plots by doing the following: set switch 18 ON to suppress the standard plot titles generate a plot with the node data outside EMME/2, transform the plot into an annotation (annotx) using GPL and the ANNOT.GPL format file (gpl -f annot.gpl -o annotx plots) generate a plot with the link data and recall annotation x (using the previous window). 5.2.6: Why is it that, when I use link color index in Module 6.12, I do not see all my links? Check that the "missing" links have a positive color index. Except in graphic worksheets, all links with a color index less than or equal to 0 are not drawn or drawn in the background color. 5.2.7: Can the information added to the plot with interactive commands in Modules 2.15, 2.43 and 6.15 be written to the plot file? Normally, this information is not sent to the plot file. However, if switch 14 is ON, all graphic output which is transmitted to the terminal is also copied to the plot file, using the terminal device definition. Note that this mode of operation is not standard and is NOT recommended for general use since it creates plot files that may be difficult to process with GPL, GPR or E2PLOTS. You can also use the Print Page option in EMTOOLW to print all the information actually on the screen. 5.2.8: How can I obtain a plot where node numbers (labels) are displayed for centroids only? Starting with Release 9, you can use the new sub-option 3 (several user defined link attributes) of option 6 (user defined data) to customize the display of link attributes, together with the specification of the subset of nodes for which the node number or label is displayed, by choosing option 3 of the following new Select question: Select: Type of node number display 1= display no node numbers 2= display all node numbers 3= display selected node numbers For more details, refer to the EMME/2 User's Manual, Module 2.13, option 6. For releases prior to Release 9, the cent2ann.awk script, available in the EMME/2, STAN Scripts section, can be used to generate an annotation file containing only the centroid numbers or labels, using the network input file (d211.in) as input. 5.3 Network and Matrix Calculators 5.3.1: How can I set color indices for different types of links? The best way is to use the network calculator (Module 2.41) to set the values in a link attribute. For example, if you want to color the connectors in green, the expression would simply be 3 (for color index 3) and, to apply the result to the connectors only, the link selection would be: ci=1 |cj=1 Note that the link attribute must first be initialized to the desired (positive) color for all other links since links with a color index of 0 are not displayed. 5.3.2: Can temporary link attributes tmplx be used in selections? No. However, it is possible to include conditions within the expression in order to "simulate" the selection. For example, if you want to change the length of the links only for those links having tmpl1 greater than 0, the expression would be: len * (tmpl1 <= 0) + expression * (tmpl1 > 0) It is also possible to replace temporary attributes by extra attributes, which can be used in the selections. 5.3.3: How can I write an expression that ensures a negative value is replaced by 0? Use the .max. operator: (expression) .max.0. 5.3.4: In the network and matrix calculators, why are the results saved when I send the report to the printer but not when I send the report to the terminal? Probably because you stopped the display of the report prematurely with a r or q. If you do not want to look at the remainder of the report, type l to skip to the end of the report. 5.4 Matrices 5.4.1: Why do some of my matrices have a "/c" flag? Probably because a transit assignment did not end normally. During a transit assignment, matrices are transformed to columnwise format in order to work destination by destination. At the end of the assignment, the matrices are retransformed to rowwise format. If the transit assignment does not end normally, the matrices will stay columnwise. Since working with columnwise matrices will increase the matrix access time drastically, you should change the storage type of columnwise matrices by using Module 3.12, option 3 (modify a matrix). 5.4.2: Matrix calculations are very slow. Is there a way to speed this up? You could do the following: Check that the matrices used in the calculations are not columnwise (see also Question 5.4.1). When running a macro with many calculations, suppress programmed pauses (~o|32). On Intel platforms, try to perform disk defragmentation. If you are using DOS or Windows 3.x, verify the size of your disk cache (Smartdrive). On UNIX platforms, try to use the data bank locally and not via a network. 5.4.3: How can I list several origin and destination matrices, excluding the zones for which the value is 0 for all matrices? Prepare an origin matrix that will be used as a constraint matrix. This matrix should contain the sum of all origin and destination matrices (destination matrices should be transposed!). For example: mo99 = mo1 + mo2 + md3'. If matrices contain negative values, the expression should be: abs(mo1) + abs(mo2) + abs(md3'). 5.5 Functions 5.5.1: How can I try different values for a constant that appears in my volume-delay function without entering the complete function each time? Put the value in a scalar matrix n and then refer to it in the function using ms(n). 5.5.2: One of the sample data files provided with the EMME/2 system contains functions. Are they general enough to be used with any data bank? No. These functions are specific to the Winnipeg data and are included solely for demonstration purposes. 5.6 Assignments 5.6.1: After an auto assignment, are the times on the connectors included in the auto time saved as a matrix? Yes. The time will depend on the functions for the connectors. 5.6.2: Is it possible to stop an auto assignment before one of the specified stopping criteria is met? Yes. Use Ctrl-C. The assignment will stop at the end of the next iteration. 5.6.3: How can I identify the non-assigned trips after an assignment? You can output the demand matrix using the time matrix as a constraint matrix with the constraint interval: (0.001,9999,exclude). This will include the trips on the diagonal (travel time of zero) and the non-connected trips (travel time equal to infinity). 5.6.4: I have a travel time matrix. How can I obtain the minimum travel time for each origin? In Module 3.21, perform a matrix computation and save the result in an origin matrix. Use the travel time matrix for the expression and as a constraint matrix (to exclude 0 values). Use the .min. operator for aggregation over destinations. 5.6.5: How can I perform an additional options auto assignment AND consider transit vehicles as additional volumes at the same time? You can compute the auto equivalent of the transit vehicles in Module 2.41 (vauteq*60/hdwy) and store the results in link and turn user data. Then, replace volad by ulN in the volume-delay functions, and pvolad by upN in the penalty functions. 5.6.6: How can I optimize the speed of my assignment? You can suppress the computation of the objective function by setting the relative gap to a negative value (for example, -1). You should also simplify your function expressions as much as possible. For example, replace sub-expressions such as constantA/constantB by constantC and use put() and get() functions whenever possible. 5.7 Zone Systems 5.7.1: Can I use different zone systems in the same data bank? It is possible but not recommended. It is better to have different data banks for different zone systems. Refer to the EMME/2 User's Manual, Chapter 3, Section 3.2.2 Matrices and scenarios. 5.7.2: I need more detailed zones in some parts of my network and more aggregate data in other parts. What is the best way to modify my zone system? Refer to the EMME/2 News of March 87 - Moving matrix data between different zone systems. The basic idea is to: create 2 data banks, 1 for each zone system in each data bank, create an ensemble that will be used to define the correspondence between the 2 zone systems punch and read the matrices using this ensemble (further manipulations might be required) 5.7.3: How can I increase the number of centroids in my data bank? While it is possible to increase most of the data bank dimensions by using the macro modifydb, it is not possible to increase the number of centroids. If this is necessary, the only solution is to create a new data bank with the new dimensions. Data from the previous data bank can then be imported using Module 1.31.
This is because the new file was not read into the data bank, either because switch 10 was not set, or because the wrong file was read. In order to input the contents of the new device table file, perform the following:
The new device table should be named d002.in and should be in the current application directory. If this is NOT the case, set the file name using the menu command batchin=? and specify the correct file name, including the complete path if the file is not in the current application directory.
It is possible but the window coordinates shown on network plots may be displayed as *******. Thirteen characters are allocated to display the window coordinates, with a maximum of 7 characters for each number, including any necessary decimal points and minus signs. A coordinate with decimals will be rounded to fit the 7 character field width, if possible. Any coordinate which cannot be displayed using 7 characters, for example a positive integer with more than 7 digits, will always be displayed as *******.
5.0.5: Before leaving a module I tried to look at the batchout file using for example, "~! type batchout", but the file was empty. Why? The batchout file is closed only when leaving the module. Before leaving the module, the contents of the batchout file may be incomplete because the file buffer is only emptied when the file is closed. Note that the same is true for the reports, plots and errors files. Therefore, always quit a module before trying to use operating system commands for file access via the command escape (~ !). 5.1 Network Modifications 5.1.1: Why do I get errors when trying to read a d211.in file containing node numbers with more than 4 digits? By default, the number of digits for node numbers is limited to 4. However, it is possible to use node numbers of up to 6 digits. To change the default, use option 7 of Module 1.23. (See also Question 5.0.2). 5.1.2: The length of my links was entered using *. Will the length be updated automatically if I modify the coordinates of a node? No. You will have to modify the attribute len of the links using that node. This can be done in Module 2.12, in Module 2.11 or in Module 2.41, where the Euclidean distance between the I- and J-node can be computed using the expression: sqrt((xi-xj)^2+(yi-yj)^2)* scale factor (the scale factor is the length of a coordinate unit as specified in Module 1.23). The links can be selected using: i=node number or j=node number 5.1.3: How can I modify a link attribute? There are many possibilities. If you have only a few links to modify, you can use the graphic worksheet (Module 2.12). You can punch the links to be modified (Module 2.14), edit the batchout file and read the new file in batch (Module 2.11). If the new values are contained in a file which is NOT in d211.in format, use option 3 of Module 2.41 to input the values in batch. If the modifications are systematic, you can use Module 2.41 and save the result of the expression (which can be a constant) in an attribute. In the link selection, be sure to include only the links to be modified. 5.1.4: I would like to initialize my base network and read a new d211.in file but I am not allowed to because there are turns (and/or transit lines) already defined. How can I read my file without losing my turns (and/or transit lines)? This is not possible due to the hierarchy of the data bank: modes (and vehicles), nodes, links, turns, transit lines (see the EMME/2 User's Manual, Chapter 3, Figure 3-1). You should thus proceed as follows. Punch all turns using Module 2.31, then delete all turns. Punch all transit lines using Module 2.24, delete all transit lines. Initialize the network, read the new d211.in file and read back the turn and transit line batchout files. Turns and transit lines which are not compatible with the new network will generate errors. 5.2 Network Plots 5.2.1: How do I change the size of arrowheads on network plots? The size of the arrowheads can be modified in Module 1.23. The default size is 2.0. Using 0 will give links with no arrowheads. 5.2.2: How can I obtain a plot where nodes are connected by only one vector (even if two-way links exist) and with no arrowhead? Use Module 1.23 to set the size of the arrowheads to 0. In the desired modules, set the module parameter Offset to the right for links to 0. 5.2.3: Once I have a plot, how can I zoom out? Use the graphic command w followed by a number between 2 and 9. This will result in a window centered on the same node but covering an area 2 to 9 times larger. 5.2.4: In Module 2.13, how can I display the value of link or node attribute with a fixed number of decimals? You can create a new extra attribute and compute the value with the required number of decimals in Module 2.41. For example, to display the value of ul3 with 2 decimals, perform the following calculation: @att=nint(ul3*100)/100 and then, display the value of @att in Module 2.13. Starting with Release 9, you can also use the new sub-option 3 (several user defined link attributes) of option 6 (user defined data) to customize the display of link attributes, including the specification of the number of decimals. When selecting this option, you can fix the number of digits after the decimal point by specifying the selected number (y) as the second value for the Enter: Color index for layer N question following the specification of the attribute (see the EMME/2 User's Manual, Module 2.13, option 6). 5.2.5: Is it possible to display node AND link data on the same plot? No. However, it is possible to superimpose 2 plots by doing the following: set switch 18 ON to suppress the standard plot titles generate a plot with the node data outside EMME/2, transform the plot into an annotation (annotx) using GPL and the ANNOT.GPL format file (gpl -f annot.gpl -o annotx plots) generate a plot with the link data and recall annotation x (using the previous window). 5.2.6: Why is it that, when I use link color index in Module 6.12, I do not see all my links? Check that the "missing" links have a positive color index. Except in graphic worksheets, all links with a color index less than or equal to 0 are not drawn or drawn in the background color. 5.2.7: Can the information added to the plot with interactive commands in Modules 2.15, 2.43 and 6.15 be written to the plot file? Normally, this information is not sent to the plot file. However, if switch 14 is ON, all graphic output which is transmitted to the terminal is also copied to the plot file, using the terminal device definition. Note that this mode of operation is not standard and is NOT recommended for general use since it creates plot files that may be difficult to process with GPL, GPR or E2PLOTS. You can also use the Print Page option in EMTOOLW to print all the information actually on the screen. 5.2.8: How can I obtain a plot where node numbers (labels) are displayed for centroids only? Starting with Release 9, you can use the new sub-option 3 (several user defined link attributes) of option 6 (user defined data) to customize the display of link attributes, together with the specification of the subset of nodes for which the node number or label is displayed, by choosing option 3 of the following new Select question: Select: Type of node number display 1= display no node numbers 2= display all node numbers 3= display selected node numbers For more details, refer to the EMME/2 User's Manual, Module 2.13, option 6. For releases prior to Release 9, the cent2ann.awk script, available in the EMME/2, STAN Scripts section, can be used to generate an annotation file containing only the centroid numbers or labels, using the network input file (d211.in) as input. 5.3 Network and Matrix Calculators 5.3.1: How can I set color indices for different types of links? The best way is to use the network calculator (Module 2.41) to set the values in a link attribute. For example, if you want to color the connectors in green, the expression would simply be 3 (for color index 3) and, to apply the result to the connectors only, the link selection would be: ci=1 |cj=1 Note that the link attribute must first be initialized to the desired (positive) color for all other links since links with a color index of 0 are not displayed. 5.3.2: Can temporary link attributes tmplx be used in selections? No. However, it is possible to include conditions within the expression in order to "simulate" the selection. For example, if you want to change the length of the links only for those links having tmpl1 greater than 0, the expression would be: len * (tmpl1 <= 0) + expression * (tmpl1 > 0) It is also possible to replace temporary attributes by extra attributes, which can be used in the selections. 5.3.3: How can I write an expression that ensures a negative value is replaced by 0? Use the .max. operator: (expression) .max.0. 5.3.4: In the network and matrix calculators, why are the results saved when I send the report to the printer but not when I send the report to the terminal? Probably because you stopped the display of the report prematurely with a r or q. If you do not want to look at the remainder of the report, type l to skip to the end of the report. 5.4 Matrices 5.4.1: Why do some of my matrices have a "/c" flag? Probably because a transit assignment did not end normally. During a transit assignment, matrices are transformed to columnwise format in order to work destination by destination. At the end of the assignment, the matrices are retransformed to rowwise format. If the transit assignment does not end normally, the matrices will stay columnwise. Since working with columnwise matrices will increase the matrix access time drastically, you should change the storage type of columnwise matrices by using Module 3.12, option 3 (modify a matrix). 5.4.2: Matrix calculations are very slow. Is there a way to speed this up? You could do the following: Check that the matrices used in the calculations are not columnwise (see also Question 5.4.1). When running a macro with many calculations, suppress programmed pauses (~o|32). On Intel platforms, try to perform disk defragmentation. If you are using DOS or Windows 3.x, verify the size of your disk cache (Smartdrive). On UNIX platforms, try to use the data bank locally and not via a network. 5.4.3: How can I list several origin and destination matrices, excluding the zones for which the value is 0 for all matrices? Prepare an origin matrix that will be used as a constraint matrix. This matrix should contain the sum of all origin and destination matrices (destination matrices should be transposed!). For example: mo99 = mo1 + mo2 + md3'. If matrices contain negative values, the expression should be: abs(mo1) + abs(mo2) + abs(md3'). 5.5 Functions 5.5.1: How can I try different values for a constant that appears in my volume-delay function without entering the complete function each time? Put the value in a scalar matrix n and then refer to it in the function using ms(n). 5.5.2: One of the sample data files provided with the EMME/2 system contains functions. Are they general enough to be used with any data bank? No. These functions are specific to the Winnipeg data and are included solely for demonstration purposes. 5.6 Assignments 5.6.1: After an auto assignment, are the times on the connectors included in the auto time saved as a matrix? Yes. The time will depend on the functions for the connectors. 5.6.2: Is it possible to stop an auto assignment before one of the specified stopping criteria is met? Yes. Use Ctrl-C. The assignment will stop at the end of the next iteration. 5.6.3: How can I identify the non-assigned trips after an assignment? You can output the demand matrix using the time matrix as a constraint matrix with the constraint interval: (0.001,9999,exclude). This will include the trips on the diagonal (travel time of zero) and the non-connected trips (travel time equal to infinity). 5.6.4: I have a travel time matrix. How can I obtain the minimum travel time for each origin? In Module 3.21, perform a matrix computation and save the result in an origin matrix. Use the travel time matrix for the expression and as a constraint matrix (to exclude 0 values). Use the .min. operator for aggregation over destinations. 5.6.5: How can I perform an additional options auto assignment AND consider transit vehicles as additional volumes at the same time? You can compute the auto equivalent of the transit vehicles in Module 2.41 (vauteq*60/hdwy) and store the results in link and turn user data. Then, replace volad by ulN in the volume-delay functions, and pvolad by upN in the penalty functions. 5.6.6: How can I optimize the speed of my assignment? You can suppress the computation of the objective function by setting the relative gap to a negative value (for example, -1). You should also simplify your function expressions as much as possible. For example, replace sub-expressions such as constantA/constantB by constantC and use put() and get() functions whenever possible. 5.7 Zone Systems 5.7.1: Can I use different zone systems in the same data bank? It is possible but not recommended. It is better to have different data banks for different zone systems. Refer to the EMME/2 User's Manual, Chapter 3, Section 3.2.2 Matrices and scenarios. 5.7.2: I need more detailed zones in some parts of my network and more aggregate data in other parts. What is the best way to modify my zone system? Refer to the EMME/2 News of March 87 - Moving matrix data between different zone systems. The basic idea is to: create 2 data banks, 1 for each zone system in each data bank, create an ensemble that will be used to define the correspondence between the 2 zone systems punch and read the matrices using this ensemble (further manipulations might be required) 5.7.3: How can I increase the number of centroids in my data bank? While it is possible to increase most of the data bank dimensions by using the macro modifydb, it is not possible to increase the number of centroids. If this is necessary, the only solution is to create a new data bank with the new dimensions. Data from the previous data bank can then be imported using Module 1.31.
The batchout file is closed only when leaving the module. Before leaving the module, the contents of the batchout file may be incomplete because the file buffer is only emptied when the file is closed. Note that the same is true for the reports, plots and errors files. Therefore, always quit a module before trying to use operating system commands for file access via the command escape (~ !).
5.1 Network Modifications 5.1.1: Why do I get errors when trying to read a d211.in file containing node numbers with more than 4 digits? By default, the number of digits for node numbers is limited to 4. However, it is possible to use node numbers of up to 6 digits. To change the default, use option 7 of Module 1.23. (See also Question 5.0.2). 5.1.2: The length of my links was entered using *. Will the length be updated automatically if I modify the coordinates of a node? No. You will have to modify the attribute len of the links using that node. This can be done in Module 2.12, in Module 2.11 or in Module 2.41, where the Euclidean distance between the I- and J-node can be computed using the expression: sqrt((xi-xj)^2+(yi-yj)^2)* scale factor (the scale factor is the length of a coordinate unit as specified in Module 1.23). The links can be selected using: i=node number or j=node number 5.1.3: How can I modify a link attribute? There are many possibilities. If you have only a few links to modify, you can use the graphic worksheet (Module 2.12). You can punch the links to be modified (Module 2.14), edit the batchout file and read the new file in batch (Module 2.11). If the new values are contained in a file which is NOT in d211.in format, use option 3 of Module 2.41 to input the values in batch. If the modifications are systematic, you can use Module 2.41 and save the result of the expression (which can be a constant) in an attribute. In the link selection, be sure to include only the links to be modified. 5.1.4: I would like to initialize my base network and read a new d211.in file but I am not allowed to because there are turns (and/or transit lines) already defined. How can I read my file without losing my turns (and/or transit lines)? This is not possible due to the hierarchy of the data bank: modes (and vehicles), nodes, links, turns, transit lines (see the EMME/2 User's Manual, Chapter 3, Figure 3-1). You should thus proceed as follows. Punch all turns using Module 2.31, then delete all turns. Punch all transit lines using Module 2.24, delete all transit lines. Initialize the network, read the new d211.in file and read back the turn and transit line batchout files. Turns and transit lines which are not compatible with the new network will generate errors. 5.2 Network Plots 5.2.1: How do I change the size of arrowheads on network plots? The size of the arrowheads can be modified in Module 1.23. The default size is 2.0. Using 0 will give links with no arrowheads. 5.2.2: How can I obtain a plot where nodes are connected by only one vector (even if two-way links exist) and with no arrowhead? Use Module 1.23 to set the size of the arrowheads to 0. In the desired modules, set the module parameter Offset to the right for links to 0. 5.2.3: Once I have a plot, how can I zoom out? Use the graphic command w followed by a number between 2 and 9. This will result in a window centered on the same node but covering an area 2 to 9 times larger. 5.2.4: In Module 2.13, how can I display the value of link or node attribute with a fixed number of decimals? You can create a new extra attribute and compute the value with the required number of decimals in Module 2.41. For example, to display the value of ul3 with 2 decimals, perform the following calculation: @att=nint(ul3*100)/100 and then, display the value of @att in Module 2.13. Starting with Release 9, you can also use the new sub-option 3 (several user defined link attributes) of option 6 (user defined data) to customize the display of link attributes, including the specification of the number of decimals. When selecting this option, you can fix the number of digits after the decimal point by specifying the selected number (y) as the second value for the Enter: Color index for layer N question following the specification of the attribute (see the EMME/2 User's Manual, Module 2.13, option 6). 5.2.5: Is it possible to display node AND link data on the same plot? No. However, it is possible to superimpose 2 plots by doing the following: set switch 18 ON to suppress the standard plot titles generate a plot with the node data outside EMME/2, transform the plot into an annotation (annotx) using GPL and the ANNOT.GPL format file (gpl -f annot.gpl -o annotx plots) generate a plot with the link data and recall annotation x (using the previous window). 5.2.6: Why is it that, when I use link color index in Module 6.12, I do not see all my links? Check that the "missing" links have a positive color index. Except in graphic worksheets, all links with a color index less than or equal to 0 are not drawn or drawn in the background color. 5.2.7: Can the information added to the plot with interactive commands in Modules 2.15, 2.43 and 6.15 be written to the plot file? Normally, this information is not sent to the plot file. However, if switch 14 is ON, all graphic output which is transmitted to the terminal is also copied to the plot file, using the terminal device definition. Note that this mode of operation is not standard and is NOT recommended for general use since it creates plot files that may be difficult to process with GPL, GPR or E2PLOTS. You can also use the Print Page option in EMTOOLW to print all the information actually on the screen. 5.2.8: How can I obtain a plot where node numbers (labels) are displayed for centroids only? Starting with Release 9, you can use the new sub-option 3 (several user defined link attributes) of option 6 (user defined data) to customize the display of link attributes, together with the specification of the subset of nodes for which the node number or label is displayed, by choosing option 3 of the following new Select question: Select: Type of node number display 1= display no node numbers 2= display all node numbers 3= display selected node numbers For more details, refer to the EMME/2 User's Manual, Module 2.13, option 6. For releases prior to Release 9, the cent2ann.awk script, available in the EMME/2, STAN Scripts section, can be used to generate an annotation file containing only the centroid numbers or labels, using the network input file (d211.in) as input. 5.3 Network and Matrix Calculators 5.3.1: How can I set color indices for different types of links? The best way is to use the network calculator (Module 2.41) to set the values in a link attribute. For example, if you want to color the connectors in green, the expression would simply be 3 (for color index 3) and, to apply the result to the connectors only, the link selection would be: ci=1 |cj=1 Note that the link attribute must first be initialized to the desired (positive) color for all other links since links with a color index of 0 are not displayed. 5.3.2: Can temporary link attributes tmplx be used in selections? No. However, it is possible to include conditions within the expression in order to "simulate" the selection. For example, if you want to change the length of the links only for those links having tmpl1 greater than 0, the expression would be: len * (tmpl1 <= 0) + expression * (tmpl1 > 0) It is also possible to replace temporary attributes by extra attributes, which can be used in the selections. 5.3.3: How can I write an expression that ensures a negative value is replaced by 0? Use the .max. operator: (expression) .max.0. 5.3.4: In the network and matrix calculators, why are the results saved when I send the report to the printer but not when I send the report to the terminal? Probably because you stopped the display of the report prematurely with a r or q. If you do not want to look at the remainder of the report, type l to skip to the end of the report. 5.4 Matrices 5.4.1: Why do some of my matrices have a "/c" flag? Probably because a transit assignment did not end normally. During a transit assignment, matrices are transformed to columnwise format in order to work destination by destination. At the end of the assignment, the matrices are retransformed to rowwise format. If the transit assignment does not end normally, the matrices will stay columnwise. Since working with columnwise matrices will increase the matrix access time drastically, you should change the storage type of columnwise matrices by using Module 3.12, option 3 (modify a matrix). 5.4.2: Matrix calculations are very slow. Is there a way to speed this up? You could do the following: Check that the matrices used in the calculations are not columnwise (see also Question 5.4.1). When running a macro with many calculations, suppress programmed pauses (~o|32). On Intel platforms, try to perform disk defragmentation. If you are using DOS or Windows 3.x, verify the size of your disk cache (Smartdrive). On UNIX platforms, try to use the data bank locally and not via a network. 5.4.3: How can I list several origin and destination matrices, excluding the zones for which the value is 0 for all matrices? Prepare an origin matrix that will be used as a constraint matrix. This matrix should contain the sum of all origin and destination matrices (destination matrices should be transposed!). For example: mo99 = mo1 + mo2 + md3'. If matrices contain negative values, the expression should be: abs(mo1) + abs(mo2) + abs(md3'). 5.5 Functions 5.5.1: How can I try different values for a constant that appears in my volume-delay function without entering the complete function each time? Put the value in a scalar matrix n and then refer to it in the function using ms(n). 5.5.2: One of the sample data files provided with the EMME/2 system contains functions. Are they general enough to be used with any data bank? No. These functions are specific to the Winnipeg data and are included solely for demonstration purposes. 5.6 Assignments 5.6.1: After an auto assignment, are the times on the connectors included in the auto time saved as a matrix? Yes. The time will depend on the functions for the connectors. 5.6.2: Is it possible to stop an auto assignment before one of the specified stopping criteria is met? Yes. Use Ctrl-C. The assignment will stop at the end of the next iteration. 5.6.3: How can I identify the non-assigned trips after an assignment? You can output the demand matrix using the time matrix as a constraint matrix with the constraint interval: (0.001,9999,exclude). This will include the trips on the diagonal (travel time of zero) and the non-connected trips (travel time equal to infinity). 5.6.4: I have a travel time matrix. How can I obtain the minimum travel time for each origin? In Module 3.21, perform a matrix computation and save the result in an origin matrix. Use the travel time matrix for the expression and as a constraint matrix (to exclude 0 values). Use the .min. operator for aggregation over destinations. 5.6.5: How can I perform an additional options auto assignment AND consider transit vehicles as additional volumes at the same time? You can compute the auto equivalent of the transit vehicles in Module 2.41 (vauteq*60/hdwy) and store the results in link and turn user data. Then, replace volad by ulN in the volume-delay functions, and pvolad by upN in the penalty functions. 5.6.6: How can I optimize the speed of my assignment? You can suppress the computation of the objective function by setting the relative gap to a negative value (for example, -1). You should also simplify your function expressions as much as possible. For example, replace sub-expressions such as constantA/constantB by constantC and use put() and get() functions whenever possible. 5.7 Zone Systems 5.7.1: Can I use different zone systems in the same data bank? It is possible but not recommended. It is better to have different data banks for different zone systems. Refer to the EMME/2 User's Manual, Chapter 3, Section 3.2.2 Matrices and scenarios. 5.7.2: I need more detailed zones in some parts of my network and more aggregate data in other parts. What is the best way to modify my zone system? Refer to the EMME/2 News of March 87 - Moving matrix data between different zone systems. The basic idea is to: create 2 data banks, 1 for each zone system in each data bank, create an ensemble that will be used to define the correspondence between the 2 zone systems punch and read the matrices using this ensemble (further manipulations might be required) 5.7.3: How can I increase the number of centroids in my data bank? While it is possible to increase most of the data bank dimensions by using the macro modifydb, it is not possible to increase the number of centroids. If this is necessary, the only solution is to create a new data bank with the new dimensions. Data from the previous data bank can then be imported using Module 1.31.
5.1.1: Why do I get errors when trying to read a d211.in file containing node numbers with more than 4 digits? By default, the number of digits for node numbers is limited to 4. However, it is possible to use node numbers of up to 6 digits. To change the default, use option 7 of Module 1.23. (See also Question 5.0.2). 5.1.2: The length of my links was entered using *. Will the length be updated automatically if I modify the coordinates of a node? No. You will have to modify the attribute len of the links using that node. This can be done in Module 2.12, in Module 2.11 or in Module 2.41, where the Euclidean distance between the I- and J-node can be computed using the expression: sqrt((xi-xj)^2+(yi-yj)^2)* scale factor (the scale factor is the length of a coordinate unit as specified in Module 1.23). The links can be selected using: i=node number or j=node number 5.1.3: How can I modify a link attribute? There are many possibilities. If you have only a few links to modify, you can use the graphic worksheet (Module 2.12). You can punch the links to be modified (Module 2.14), edit the batchout file and read the new file in batch (Module 2.11). If the new values are contained in a file which is NOT in d211.in format, use option 3 of Module 2.41 to input the values in batch. If the modifications are systematic, you can use Module 2.41 and save the result of the expression (which can be a constant) in an attribute. In the link selection, be sure to include only the links to be modified. 5.1.4: I would like to initialize my base network and read a new d211.in file but I am not allowed to because there are turns (and/or transit lines) already defined. How can I read my file without losing my turns (and/or transit lines)? This is not possible due to the hierarchy of the data bank: modes (and vehicles), nodes, links, turns, transit lines (see the EMME/2 User's Manual, Chapter 3, Figure 3-1). You should thus proceed as follows. Punch all turns using Module 2.31, then delete all turns. Punch all transit lines using Module 2.24, delete all transit lines. Initialize the network, read the new d211.in file and read back the turn and transit line batchout files. Turns and transit lines which are not compatible with the new network will generate errors. 5.2 Network Plots 5.2.1: How do I change the size of arrowheads on network plots? The size of the arrowheads can be modified in Module 1.23. The default size is 2.0. Using 0 will give links with no arrowheads. 5.2.2: How can I obtain a plot where nodes are connected by only one vector (even if two-way links exist) and with no arrowhead? Use Module 1.23 to set the size of the arrowheads to 0. In the desired modules, set the module parameter Offset to the right for links to 0. 5.2.3: Once I have a plot, how can I zoom out? Use the graphic command w followed by a number between 2 and 9. This will result in a window centered on the same node but covering an area 2 to 9 times larger. 5.2.4: In Module 2.13, how can I display the value of link or node attribute with a fixed number of decimals? You can create a new extra attribute and compute the value with the required number of decimals in Module 2.41. For example, to display the value of ul3 with 2 decimals, perform the following calculation: @att=nint(ul3*100)/100 and then, display the value of @att in Module 2.13. Starting with Release 9, you can also use the new sub-option 3 (several user defined link attributes) of option 6 (user defined data) to customize the display of link attributes, including the specification of the number of decimals. When selecting this option, you can fix the number of digits after the decimal point by specifying the selected number (y) as the second value for the Enter: Color index for layer N question following the specification of the attribute (see the EMME/2 User's Manual, Module 2.13, option 6). 5.2.5: Is it possible to display node AND link data on the same plot? No. However, it is possible to superimpose 2 plots by doing the following: set switch 18 ON to suppress the standard plot titles generate a plot with the node data outside EMME/2, transform the plot into an annotation (annotx) using GPL and the ANNOT.GPL format file (gpl -f annot.gpl -o annotx plots) generate a plot with the link data and recall annotation x (using the previous window). 5.2.6: Why is it that, when I use link color index in Module 6.12, I do not see all my links? Check that the "missing" links have a positive color index. Except in graphic worksheets, all links with a color index less than or equal to 0 are not drawn or drawn in the background color. 5.2.7: Can the information added to the plot with interactive commands in Modules 2.15, 2.43 and 6.15 be written to the plot file? Normally, this information is not sent to the plot file. However, if switch 14 is ON, all graphic output which is transmitted to the terminal is also copied to the plot file, using the terminal device definition. Note that this mode of operation is not standard and is NOT recommended for general use since it creates plot files that may be difficult to process with GPL, GPR or E2PLOTS. You can also use the Print Page option in EMTOOLW to print all the information actually on the screen. 5.2.8: How can I obtain a plot where node numbers (labels) are displayed for centroids only? Starting with Release 9, you can use the new sub-option 3 (several user defined link attributes) of option 6 (user defined data) to customize the display of link attributes, together with the specification of the subset of nodes for which the node number or label is displayed, by choosing option 3 of the following new Select question: Select: Type of node number display 1= display no node numbers 2= display all node numbers 3= display selected node numbers For more details, refer to the EMME/2 User's Manual, Module 2.13, option 6. For releases prior to Release 9, the cent2ann.awk script, available in the EMME/2, STAN Scripts section, can be used to generate an annotation file containing only the centroid numbers or labels, using the network input file (d211.in) as input. 5.3 Network and Matrix Calculators 5.3.1: How can I set color indices for different types of links? The best way is to use the network calculator (Module 2.41) to set the values in a link attribute. For example, if you want to color the connectors in green, the expression would simply be 3 (for color index 3) and, to apply the result to the connectors only, the link selection would be: ci=1 |cj=1 Note that the link attribute must first be initialized to the desired (positive) color for all other links since links with a color index of 0 are not displayed. 5.3.2: Can temporary link attributes tmplx be used in selections? No. However, it is possible to include conditions within the expression in order to "simulate" the selection. For example, if you want to change the length of the links only for those links having tmpl1 greater than 0, the expression would be: len * (tmpl1 <= 0) + expression * (tmpl1 > 0) It is also possible to replace temporary attributes by extra attributes, which can be used in the selections. 5.3.3: How can I write an expression that ensures a negative value is replaced by 0? Use the .max. operator: (expression) .max.0. 5.3.4: In the network and matrix calculators, why are the results saved when I send the report to the printer but not when I send the report to the terminal? Probably because you stopped the display of the report prematurely with a r or q. If you do not want to look at the remainder of the report, type l to skip to the end of the report. 5.4 Matrices 5.4.1: Why do some of my matrices have a "/c" flag? Probably because a transit assignment did not end normally. During a transit assignment, matrices are transformed to columnwise format in order to work destination by destination. At the end of the assignment, the matrices are retransformed to rowwise format. If the transit assignment does not end normally, the matrices will stay columnwise. Since working with columnwise matrices will increase the matrix access time drastically, you should change the storage type of columnwise matrices by using Module 3.12, option 3 (modify a matrix). 5.4.2: Matrix calculations are very slow. Is there a way to speed this up? You could do the following: Check that the matrices used in the calculations are not columnwise (see also Question 5.4.1). When running a macro with many calculations, suppress programmed pauses (~o|32). On Intel platforms, try to perform disk defragmentation. If you are using DOS or Windows 3.x, verify the size of your disk cache (Smartdrive). On UNIX platforms, try to use the data bank locally and not via a network. 5.4.3: How can I list several origin and destination matrices, excluding the zones for which the value is 0 for all matrices? Prepare an origin matrix that will be used as a constraint matrix. This matrix should contain the sum of all origin and destination matrices (destination matrices should be transposed!). For example: mo99 = mo1 + mo2 + md3'. If matrices contain negative values, the expression should be: abs(mo1) + abs(mo2) + abs(md3'). 5.5 Functions 5.5.1: How can I try different values for a constant that appears in my volume-delay function without entering the complete function each time? Put the value in a scalar matrix n and then refer to it in the function using ms(n). 5.5.2: One of the sample data files provided with the EMME/2 system contains functions. Are they general enough to be used with any data bank? No. These functions are specific to the Winnipeg data and are included solely for demonstration purposes. 5.6 Assignments 5.6.1: After an auto assignment, are the times on the connectors included in the auto time saved as a matrix? Yes. The time will depend on the functions for the connectors. 5.6.2: Is it possible to stop an auto assignment before one of the specified stopping criteria is met? Yes. Use Ctrl-C. The assignment will stop at the end of the next iteration. 5.6.3: How can I identify the non-assigned trips after an assignment? You can output the demand matrix using the time matrix as a constraint matrix with the constraint interval: (0.001,9999,exclude). This will include the trips on the diagonal (travel time of zero) and the non-connected trips (travel time equal to infinity). 5.6.4: I have a travel time matrix. How can I obtain the minimum travel time for each origin? In Module 3.21, perform a matrix computation and save the result in an origin matrix. Use the travel time matrix for the expression and as a constraint matrix (to exclude 0 values). Use the .min. operator for aggregation over destinations. 5.6.5: How can I perform an additional options auto assignment AND consider transit vehicles as additional volumes at the same time? You can compute the auto equivalent of the transit vehicles in Module 2.41 (vauteq*60/hdwy) and store the results in link and turn user data. Then, replace volad by ulN in the volume-delay functions, and pvolad by upN in the penalty functions. 5.6.6: How can I optimize the speed of my assignment? You can suppress the computation of the objective function by setting the relative gap to a negative value (for example, -1). You should also simplify your function expressions as much as possible. For example, replace sub-expressions such as constantA/constantB by constantC and use put() and get() functions whenever possible. 5.7 Zone Systems 5.7.1: Can I use different zone systems in the same data bank? It is possible but not recommended. It is better to have different data banks for different zone systems. Refer to the EMME/2 User's Manual, Chapter 3, Section 3.2.2 Matrices and scenarios. 5.7.2: I need more detailed zones in some parts of my network and more aggregate data in other parts. What is the best way to modify my zone system? Refer to the EMME/2 News of March 87 - Moving matrix data between different zone systems. The basic idea is to: create 2 data banks, 1 for each zone system in each data bank, create an ensemble that will be used to define the correspondence between the 2 zone systems punch and read the matrices using this ensemble (further manipulations might be required) 5.7.3: How can I increase the number of centroids in my data bank? While it is possible to increase most of the data bank dimensions by using the macro modifydb, it is not possible to increase the number of centroids. If this is necessary, the only solution is to create a new data bank with the new dimensions. Data from the previous data bank can then be imported using Module 1.31.
By default, the number of digits for node numbers is limited to 4. However, it is possible to use node numbers of up to 6 digits. To change the default, use option 7 of Module 1.23. (See also Question 5.0.2).
5.1.2: The length of my links was entered using *. Will the length be updated automatically if I modify the coordinates of a node? No. You will have to modify the attribute len of the links using that node. This can be done in Module 2.12, in Module 2.11 or in Module 2.41, where the Euclidean distance between the I- and J-node can be computed using the expression: sqrt((xi-xj)^2+(yi-yj)^2)* scale factor (the scale factor is the length of a coordinate unit as specified in Module 1.23). The links can be selected using: i=node number or j=node number 5.1.3: How can I modify a link attribute? There are many possibilities. If you have only a few links to modify, you can use the graphic worksheet (Module 2.12). You can punch the links to be modified (Module 2.14), edit the batchout file and read the new file in batch (Module 2.11). If the new values are contained in a file which is NOT in d211.in format, use option 3 of Module 2.41 to input the values in batch. If the modifications are systematic, you can use Module 2.41 and save the result of the expression (which can be a constant) in an attribute. In the link selection, be sure to include only the links to be modified. 5.1.4: I would like to initialize my base network and read a new d211.in file but I am not allowed to because there are turns (and/or transit lines) already defined. How can I read my file without losing my turns (and/or transit lines)? This is not possible due to the hierarchy of the data bank: modes (and vehicles), nodes, links, turns, transit lines (see the EMME/2 User's Manual, Chapter 3, Figure 3-1). You should thus proceed as follows. Punch all turns using Module 2.31, then delete all turns. Punch all transit lines using Module 2.24, delete all transit lines. Initialize the network, read the new d211.in file and read back the turn and transit line batchout files. Turns and transit lines which are not compatible with the new network will generate errors. 5.2 Network Plots 5.2.1: How do I change the size of arrowheads on network plots? The size of the arrowheads can be modified in Module 1.23. The default size is 2.0. Using 0 will give links with no arrowheads. 5.2.2: How can I obtain a plot where nodes are connected by only one vector (even if two-way links exist) and with no arrowhead? Use Module 1.23 to set the size of the arrowheads to 0. In the desired modules, set the module parameter Offset to the right for links to 0. 5.2.3: Once I have a plot, how can I zoom out? Use the graphic command w followed by a number between 2 and 9. This will result in a window centered on the same node but covering an area 2 to 9 times larger. 5.2.4: In Module 2.13, how can I display the value of link or node attribute with a fixed number of decimals? You can create a new extra attribute and compute the value with the required number of decimals in Module 2.41. For example, to display the value of ul3 with 2 decimals, perform the following calculation: @att=nint(ul3*100)/100 and then, display the value of @att in Module 2.13. Starting with Release 9, you can also use the new sub-option 3 (several user defined link attributes) of option 6 (user defined data) to customize the display of link attributes, including the specification of the number of decimals. When selecting this option, you can fix the number of digits after the decimal point by specifying the selected number (y) as the second value for the Enter: Color index for layer N question following the specification of the attribute (see the EMME/2 User's Manual, Module 2.13, option 6). 5.2.5: Is it possible to display node AND link data on the same plot? No. However, it is possible to superimpose 2 plots by doing the following: set switch 18 ON to suppress the standard plot titles generate a plot with the node data outside EMME/2, transform the plot into an annotation (annotx) using GPL and the ANNOT.GPL format file (gpl -f annot.gpl -o annotx plots) generate a plot with the link data and recall annotation x (using the previous window). 5.2.6: Why is it that, when I use link color index in Module 6.12, I do not see all my links? Check that the "missing" links have a positive color index. Except in graphic worksheets, all links with a color index less than or equal to 0 are not drawn or drawn in the background color. 5.2.7: Can the information added to the plot with interactive commands in Modules 2.15, 2.43 and 6.15 be written to the plot file? Normally, this information is not sent to the plot file. However, if switch 14 is ON, all graphic output which is transmitted to the terminal is also copied to the plot file, using the terminal device definition. Note that this mode of operation is not standard and is NOT recommended for general use since it creates plot files that may be difficult to process with GPL, GPR or E2PLOTS. You can also use the Print Page option in EMTOOLW to print all the information actually on the screen. 5.2.8: How can I obtain a plot where node numbers (labels) are displayed for centroids only? Starting with Release 9, you can use the new sub-option 3 (several user defined link attributes) of option 6 (user defined data) to customize the display of link attributes, together with the specification of the subset of nodes for which the node number or label is displayed, by choosing option 3 of the following new Select question: Select: Type of node number display 1= display no node numbers 2= display all node numbers 3= display selected node numbers For more details, refer to the EMME/2 User's Manual, Module 2.13, option 6. For releases prior to Release 9, the cent2ann.awk script, available in the EMME/2, STAN Scripts section, can be used to generate an annotation file containing only the centroid numbers or labels, using the network input file (d211.in) as input. 5.3 Network and Matrix Calculators 5.3.1: How can I set color indices for different types of links? The best way is to use the network calculator (Module 2.41) to set the values in a link attribute. For example, if you want to color the connectors in green, the expression would simply be 3 (for color index 3) and, to apply the result to the connectors only, the link selection would be: ci=1 |cj=1 Note that the link attribute must first be initialized to the desired (positive) color for all other links since links with a color index of 0 are not displayed. 5.3.2: Can temporary link attributes tmplx be used in selections? No. However, it is possible to include conditions within the expression in order to "simulate" the selection. For example, if you want to change the length of the links only for those links having tmpl1 greater than 0, the expression would be: len * (tmpl1 <= 0) + expression * (tmpl1 > 0) It is also possible to replace temporary attributes by extra attributes, which can be used in the selections. 5.3.3: How can I write an expression that ensures a negative value is replaced by 0? Use the .max. operator: (expression) .max.0. 5.3.4: In the network and matrix calculators, why are the results saved when I send the report to the printer but not when I send the report to the terminal? Probably because you stopped the display of the report prematurely with a r or q. If you do not want to look at the remainder of the report, type l to skip to the end of the report. 5.4 Matrices 5.4.1: Why do some of my matrices have a "/c" flag? Probably because a transit assignment did not end normally. During a transit assignment, matrices are transformed to columnwise format in order to work destination by destination. At the end of the assignment, the matrices are retransformed to rowwise format. If the transit assignment does not end normally, the matrices will stay columnwise. Since working with columnwise matrices will increase the matrix access time drastically, you should change the storage type of columnwise matrices by using Module 3.12, option 3 (modify a matrix). 5.4.2: Matrix calculations are very slow. Is there a way to speed this up? You could do the following: Check that the matrices used in the calculations are not columnwise (see also Question 5.4.1). When running a macro with many calculations, suppress programmed pauses (~o|32). On Intel platforms, try to perform disk defragmentation. If you are using DOS or Windows 3.x, verify the size of your disk cache (Smartdrive). On UNIX platforms, try to use the data bank locally and not via a network. 5.4.3: How can I list several origin and destination matrices, excluding the zones for which the value is 0 for all matrices? Prepare an origin matrix that will be used as a constraint matrix. This matrix should contain the sum of all origin and destination matrices (destination matrices should be transposed!). For example: mo99 = mo1 + mo2 + md3'. If matrices contain negative values, the expression should be: abs(mo1) + abs(mo2) + abs(md3'). 5.5 Functions 5.5.1: How can I try different values for a constant that appears in my volume-delay function without entering the complete function each time? Put the value in a scalar matrix n and then refer to it in the function using ms(n). 5.5.2: One of the sample data files provided with the EMME/2 system contains functions. Are they general enough to be used with any data bank? No. These functions are specific to the Winnipeg data and are included solely for demonstration purposes. 5.6 Assignments 5.6.1: After an auto assignment, are the times on the connectors included in the auto time saved as a matrix? Yes. The time will depend on the functions for the connectors. 5.6.2: Is it possible to stop an auto assignment before one of the specified stopping criteria is met? Yes. Use Ctrl-C. The assignment will stop at the end of the next iteration. 5.6.3: How can I identify the non-assigned trips after an assignment? You can output the demand matrix using the time matrix as a constraint matrix with the constraint interval: (0.001,9999,exclude). This will include the trips on the diagonal (travel time of zero) and the non-connected trips (travel time equal to infinity). 5.6.4: I have a travel time matrix. How can I obtain the minimum travel time for each origin? In Module 3.21, perform a matrix computation and save the result in an origin matrix. Use the travel time matrix for the expression and as a constraint matrix (to exclude 0 values). Use the .min. operator for aggregation over destinations. 5.6.5: How can I perform an additional options auto assignment AND consider transit vehicles as additional volumes at the same time? You can compute the auto equivalent of the transit vehicles in Module 2.41 (vauteq*60/hdwy) and store the results in link and turn user data. Then, replace volad by ulN in the volume-delay functions, and pvolad by upN in the penalty functions. 5.6.6: How can I optimize the speed of my assignment? You can suppress the computation of the objective function by setting the relative gap to a negative value (for example, -1). You should also simplify your function expressions as much as possible. For example, replace sub-expressions such as constantA/constantB by constantC and use put() and get() functions whenever possible. 5.7 Zone Systems 5.7.1: Can I use different zone systems in the same data bank? It is possible but not recommended. It is better to have different data banks for different zone systems. Refer to the EMME/2 User's Manual, Chapter 3, Section 3.2.2 Matrices and scenarios. 5.7.2: I need more detailed zones in some parts of my network and more aggregate data in other parts. What is the best way to modify my zone system? Refer to the EMME/2 News of March 87 - Moving matrix data between different zone systems. The basic idea is to: create 2 data banks, 1 for each zone system in each data bank, create an ensemble that will be used to define the correspondence between the 2 zone systems punch and read the matrices using this ensemble (further manipulations might be required) 5.7.3: How can I increase the number of centroids in my data bank? While it is possible to increase most of the data bank dimensions by using the macro modifydb, it is not possible to increase the number of centroids. If this is necessary, the only solution is to create a new data bank with the new dimensions. Data from the previous data bank can then be imported using Module 1.31.
No. You will have to modify the attribute len of the links using that node. This can be done in Module 2.12, in Module 2.11 or in Module 2.41, where the Euclidean distance between the I- and J-node can be computed using the expression: sqrt((xi-xj)^2+(yi-yj)^2)* scale factor (the scale factor is the length of a coordinate unit as specified in Module 1.23). The links can be selected using:
5.1.3: How can I modify a link attribute? There are many possibilities. If you have only a few links to modify, you can use the graphic worksheet (Module 2.12). You can punch the links to be modified (Module 2.14), edit the batchout file and read the new file in batch (Module 2.11). If the new values are contained in a file which is NOT in d211.in format, use option 3 of Module 2.41 to input the values in batch. If the modifications are systematic, you can use Module 2.41 and save the result of the expression (which can be a constant) in an attribute. In the link selection, be sure to include only the links to be modified. 5.1.4: I would like to initialize my base network and read a new d211.in file but I am not allowed to because there are turns (and/or transit lines) already defined. How can I read my file without losing my turns (and/or transit lines)? This is not possible due to the hierarchy of the data bank: modes (and vehicles), nodes, links, turns, transit lines (see the EMME/2 User's Manual, Chapter 3, Figure 3-1). You should thus proceed as follows. Punch all turns using Module 2.31, then delete all turns. Punch all transit lines using Module 2.24, delete all transit lines. Initialize the network, read the new d211.in file and read back the turn and transit line batchout files. Turns and transit lines which are not compatible with the new network will generate errors. 5.2 Network Plots 5.2.1: How do I change the size of arrowheads on network plots? The size of the arrowheads can be modified in Module 1.23. The default size is 2.0. Using 0 will give links with no arrowheads. 5.2.2: How can I obtain a plot where nodes are connected by only one vector (even if two-way links exist) and with no arrowhead? Use Module 1.23 to set the size of the arrowheads to 0. In the desired modules, set the module parameter Offset to the right for links to 0. 5.2.3: Once I have a plot, how can I zoom out? Use the graphic command w followed by a number between 2 and 9. This will result in a window centered on the same node but covering an area 2 to 9 times larger. 5.2.4: In Module 2.13, how can I display the value of link or node attribute with a fixed number of decimals? You can create a new extra attribute and compute the value with the required number of decimals in Module 2.41. For example, to display the value of ul3 with 2 decimals, perform the following calculation: @att=nint(ul3*100)/100 and then, display the value of @att in Module 2.13. Starting with Release 9, you can also use the new sub-option 3 (several user defined link attributes) of option 6 (user defined data) to customize the display of link attributes, including the specification of the number of decimals. When selecting this option, you can fix the number of digits after the decimal point by specifying the selected number (y) as the second value for the Enter: Color index for layer N question following the specification of the attribute (see the EMME/2 User's Manual, Module 2.13, option 6). 5.2.5: Is it possible to display node AND link data on the same plot? No. However, it is possible to superimpose 2 plots by doing the following: set switch 18 ON to suppress the standard plot titles generate a plot with the node data outside EMME/2, transform the plot into an annotation (annotx) using GPL and the ANNOT.GPL format file (gpl -f annot.gpl -o annotx plots) generate a plot with the link data and recall annotation x (using the previous window). 5.2.6: Why is it that, when I use link color index in Module 6.12, I do not see all my links? Check that the "missing" links have a positive color index. Except in graphic worksheets, all links with a color index less than or equal to 0 are not drawn or drawn in the background color. 5.2.7: Can the information added to the plot with interactive commands in Modules 2.15, 2.43 and 6.15 be written to the plot file? Normally, this information is not sent to the plot file. However, if switch 14 is ON, all graphic output which is transmitted to the terminal is also copied to the plot file, using the terminal device definition. Note that this mode of operation is not standard and is NOT recommended for general use since it creates plot files that may be difficult to process with GPL, GPR or E2PLOTS. You can also use the Print Page option in EMTOOLW to print all the information actually on the screen. 5.2.8: How can I obtain a plot where node numbers (labels) are displayed for centroids only? Starting with Release 9, you can use the new sub-option 3 (several user defined link attributes) of option 6 (user defined data) to customize the display of link attributes, together with the specification of the subset of nodes for which the node number or label is displayed, by choosing option 3 of the following new Select question: Select: Type of node number display 1= display no node numbers 2= display all node numbers 3= display selected node numbers For more details, refer to the EMME/2 User's Manual, Module 2.13, option 6. For releases prior to Release 9, the cent2ann.awk script, available in the EMME/2, STAN Scripts section, can be used to generate an annotation file containing only the centroid numbers or labels, using the network input file (d211.in) as input. 5.3 Network and Matrix Calculators 5.3.1: How can I set color indices for different types of links? The best way is to use the network calculator (Module 2.41) to set the values in a link attribute. For example, if you want to color the connectors in green, the expression would simply be 3 (for color index 3) and, to apply the result to the connectors only, the link selection would be: ci=1 |cj=1 Note that the link attribute must first be initialized to the desired (positive) color for all other links since links with a color index of 0 are not displayed. 5.3.2: Can temporary link attributes tmplx be used in selections? No. However, it is possible to include conditions within the expression in order to "simulate" the selection. For example, if you want to change the length of the links only for those links having tmpl1 greater than 0, the expression would be: len * (tmpl1 <= 0) + expression * (tmpl1 > 0) It is also possible to replace temporary attributes by extra attributes, which can be used in the selections. 5.3.3: How can I write an expression that ensures a negative value is replaced by 0? Use the .max. operator: (expression) .max.0. 5.3.4: In the network and matrix calculators, why are the results saved when I send the report to the printer but not when I send the report to the terminal? Probably because you stopped the display of the report prematurely with a r or q. If you do not want to look at the remainder of the report, type l to skip to the end of the report. 5.4 Matrices 5.4.1: Why do some of my matrices have a "/c" flag? Probably because a transit assignment did not end normally. During a transit assignment, matrices are transformed to columnwise format in order to work destination by destination. At the end of the assignment, the matrices are retransformed to rowwise format. If the transit assignment does not end normally, the matrices will stay columnwise. Since working with columnwise matrices will increase the matrix access time drastically, you should change the storage type of columnwise matrices by using Module 3.12, option 3 (modify a matrix). 5.4.2: Matrix calculations are very slow. Is there a way to speed this up? You could do the following: Check that the matrices used in the calculations are not columnwise (see also Question 5.4.1). When running a macro with many calculations, suppress programmed pauses (~o|32). On Intel platforms, try to perform disk defragmentation. If you are using DOS or Windows 3.x, verify the size of your disk cache (Smartdrive). On UNIX platforms, try to use the data bank locally and not via a network. 5.4.3: How can I list several origin and destination matrices, excluding the zones for which the value is 0 for all matrices? Prepare an origin matrix that will be used as a constraint matrix. This matrix should contain the sum of all origin and destination matrices (destination matrices should be transposed!). For example: mo99 = mo1 + mo2 + md3'. If matrices contain negative values, the expression should be: abs(mo1) + abs(mo2) + abs(md3'). 5.5 Functions 5.5.1: How can I try different values for a constant that appears in my volume-delay function without entering the complete function each time? Put the value in a scalar matrix n and then refer to it in the function using ms(n). 5.5.2: One of the sample data files provided with the EMME/2 system contains functions. Are they general enough to be used with any data bank? No. These functions are specific to the Winnipeg data and are included solely for demonstration purposes. 5.6 Assignments 5.6.1: After an auto assignment, are the times on the connectors included in the auto time saved as a matrix? Yes. The time will depend on the functions for the connectors. 5.6.2: Is it possible to stop an auto assignment before one of the specified stopping criteria is met? Yes. Use Ctrl-C. The assignment will stop at the end of the next iteration. 5.6.3: How can I identify the non-assigned trips after an assignment? You can output the demand matrix using the time matrix as a constraint matrix with the constraint interval: (0.001,9999,exclude). This will include the trips on the diagonal (travel time of zero) and the non-connected trips (travel time equal to infinity). 5.6.4: I have a travel time matrix. How can I obtain the minimum travel time for each origin? In Module 3.21, perform a matrix computation and save the result in an origin matrix. Use the travel time matrix for the expression and as a constraint matrix (to exclude 0 values). Use the .min. operator for aggregation over destinations. 5.6.5: How can I perform an additional options auto assignment AND consider transit vehicles as additional volumes at the same time? You can compute the auto equivalent of the transit vehicles in Module 2.41 (vauteq*60/hdwy) and store the results in link and turn user data. Then, replace volad by ulN in the volume-delay functions, and pvolad by upN in the penalty functions. 5.6.6: How can I optimize the speed of my assignment? You can suppress the computation of the objective function by setting the relative gap to a negative value (for example, -1). You should also simplify your function expressions as much as possible. For example, replace sub-expressions such as constantA/constantB by constantC and use put() and get() functions whenever possible. 5.7 Zone Systems 5.7.1: Can I use different zone systems in the same data bank? It is possible but not recommended. It is better to have different data banks for different zone systems. Refer to the EMME/2 User's Manual, Chapter 3, Section 3.2.2 Matrices and scenarios. 5.7.2: I need more detailed zones in some parts of my network and more aggregate data in other parts. What is the best way to modify my zone system? Refer to the EMME/2 News of March 87 - Moving matrix data between different zone systems. The basic idea is to: create 2 data banks, 1 for each zone system in each data bank, create an ensemble that will be used to define the correspondence between the 2 zone systems punch and read the matrices using this ensemble (further manipulations might be required) 5.7.3: How can I increase the number of centroids in my data bank? While it is possible to increase most of the data bank dimensions by using the macro modifydb, it is not possible to increase the number of centroids. If this is necessary, the only solution is to create a new data bank with the new dimensions. Data from the previous data bank can then be imported using Module 1.31.
There are many possibilities.
This is not possible due to the hierarchy of the data bank: modes (and vehicles), nodes, links, turns, transit lines (see the EMME/2 User's Manual, Chapter 3, Figure 3-1). You should thus proceed as follows. Punch all turns using Module 2.31, then delete all turns. Punch all transit lines using Module 2.24, delete all transit lines. Initialize the network, read the new d211.in file and read back the turn and transit line batchout files. Turns and transit lines which are not compatible with the new network will generate errors.
5.2 Network Plots 5.2.1: How do I change the size of arrowheads on network plots? The size of the arrowheads can be modified in Module 1.23. The default size is 2.0. Using 0 will give links with no arrowheads. 5.2.2: How can I obtain a plot where nodes are connected by only one vector (even if two-way links exist) and with no arrowhead? Use Module 1.23 to set the size of the arrowheads to 0. In the desired modules, set the module parameter Offset to the right for links to 0. 5.2.3: Once I have a plot, how can I zoom out? Use the graphic command w followed by a number between 2 and 9. This will result in a window centered on the same node but covering an area 2 to 9 times larger. 5.2.4: In Module 2.13, how can I display the value of link or node attribute with a fixed number of decimals? You can create a new extra attribute and compute the value with the required number of decimals in Module 2.41. For example, to display the value of ul3 with 2 decimals, perform the following calculation: @att=nint(ul3*100)/100 and then, display the value of @att in Module 2.13. Starting with Release 9, you can also use the new sub-option 3 (several user defined link attributes) of option 6 (user defined data) to customize the display of link attributes, including the specification of the number of decimals. When selecting this option, you can fix the number of digits after the decimal point by specifying the selected number (y) as the second value for the Enter: Color index for layer N question following the specification of the attribute (see the EMME/2 User's Manual, Module 2.13, option 6). 5.2.5: Is it possible to display node AND link data on the same plot? No. However, it is possible to superimpose 2 plots by doing the following: set switch 18 ON to suppress the standard plot titles generate a plot with the node data outside EMME/2, transform the plot into an annotation (annotx) using GPL and the ANNOT.GPL format file (gpl -f annot.gpl -o annotx plots) generate a plot with the link data and recall annotation x (using the previous window). 5.2.6: Why is it that, when I use link color index in Module 6.12, I do not see all my links? Check that the "missing" links have a positive color index. Except in graphic worksheets, all links with a color index less than or equal to 0 are not drawn or drawn in the background color. 5.2.7: Can the information added to the plot with interactive commands in Modules 2.15, 2.43 and 6.15 be written to the plot file? Normally, this information is not sent to the plot file. However, if switch 14 is ON, all graphic output which is transmitted to the terminal is also copied to the plot file, using the terminal device definition. Note that this mode of operation is not standard and is NOT recommended for general use since it creates plot files that may be difficult to process with GPL, GPR or E2PLOTS. You can also use the Print Page option in EMTOOLW to print all the information actually on the screen. 5.2.8: How can I obtain a plot where node numbers (labels) are displayed for centroids only? Starting with Release 9, you can use the new sub-option 3 (several user defined link attributes) of option 6 (user defined data) to customize the display of link attributes, together with the specification of the subset of nodes for which the node number or label is displayed, by choosing option 3 of the following new Select question: Select: Type of node number display 1= display no node numbers 2= display all node numbers 3= display selected node numbers For more details, refer to the EMME/2 User's Manual, Module 2.13, option 6. For releases prior to Release 9, the cent2ann.awk script, available in the EMME/2, STAN Scripts section, can be used to generate an annotation file containing only the centroid numbers or labels, using the network input file (d211.in) as input. 5.3 Network and Matrix Calculators 5.3.1: How can I set color indices for different types of links? The best way is to use the network calculator (Module 2.41) to set the values in a link attribute. For example, if you want to color the connectors in green, the expression would simply be 3 (for color index 3) and, to apply the result to the connectors only, the link selection would be: ci=1 |cj=1 Note that the link attribute must first be initialized to the desired (positive) color for all other links since links with a color index of 0 are not displayed. 5.3.2: Can temporary link attributes tmplx be used in selections? No. However, it is possible to include conditions within the expression in order to "simulate" the selection. For example, if you want to change the length of the links only for those links having tmpl1 greater than 0, the expression would be: len * (tmpl1 <= 0) + expression * (tmpl1 > 0) It is also possible to replace temporary attributes by extra attributes, which can be used in the selections. 5.3.3: How can I write an expression that ensures a negative value is replaced by 0? Use the .max. operator: (expression) .max.0. 5.3.4: In the network and matrix calculators, why are the results saved when I send the report to the printer but not when I send the report to the terminal? Probably because you stopped the display of the report prematurely with a r or q. If you do not want to look at the remainder of the report, type l to skip to the end of the report. 5.4 Matrices 5.4.1: Why do some of my matrices have a "/c" flag? Probably because a transit assignment did not end normally. During a transit assignment, matrices are transformed to columnwise format in order to work destination by destination. At the end of the assignment, the matrices are retransformed to rowwise format. If the transit assignment does not end normally, the matrices will stay columnwise. Since working with columnwise matrices will increase the matrix access time drastically, you should change the storage type of columnwise matrices by using Module 3.12, option 3 (modify a matrix). 5.4.2: Matrix calculations are very slow. Is there a way to speed this up? You could do the following: Check that the matrices used in the calculations are not columnwise (see also Question 5.4.1). When running a macro with many calculations, suppress programmed pauses (~o|32). On Intel platforms, try to perform disk defragmentation. If you are using DOS or Windows 3.x, verify the size of your disk cache (Smartdrive). On UNIX platforms, try to use the data bank locally and not via a network. 5.4.3: How can I list several origin and destination matrices, excluding the zones for which the value is 0 for all matrices? Prepare an origin matrix that will be used as a constraint matrix. This matrix should contain the sum of all origin and destination matrices (destination matrices should be transposed!). For example: mo99 = mo1 + mo2 + md3'. If matrices contain negative values, the expression should be: abs(mo1) + abs(mo2) + abs(md3'). 5.5 Functions 5.5.1: How can I try different values for a constant that appears in my volume-delay function without entering the complete function each time? Put the value in a scalar matrix n and then refer to it in the function using ms(n). 5.5.2: One of the sample data files provided with the EMME/2 system contains functions. Are they general enough to be used with any data bank? No. These functions are specific to the Winnipeg data and are included solely for demonstration purposes. 5.6 Assignments 5.6.1: After an auto assignment, are the times on the connectors included in the auto time saved as a matrix? Yes. The time will depend on the functions for the connectors. 5.6.2: Is it possible to stop an auto assignment before one of the specified stopping criteria is met? Yes. Use Ctrl-C. The assignment will stop at the end of the next iteration. 5.6.3: How can I identify the non-assigned trips after an assignment? You can output the demand matrix using the time matrix as a constraint matrix with the constraint interval: (0.001,9999,exclude). This will include the trips on the diagonal (travel time of zero) and the non-connected trips (travel time equal to infinity). 5.6.4: I have a travel time matrix. How can I obtain the minimum travel time for each origin? In Module 3.21, perform a matrix computation and save the result in an origin matrix. Use the travel time matrix for the expression and as a constraint matrix (to exclude 0 values). Use the .min. operator for aggregation over destinations. 5.6.5: How can I perform an additional options auto assignment AND consider transit vehicles as additional volumes at the same time? You can compute the auto equivalent of the transit vehicles in Module 2.41 (vauteq*60/hdwy) and store the results in link and turn user data. Then, replace volad by ulN in the volume-delay functions, and pvolad by upN in the penalty functions. 5.6.6: How can I optimize the speed of my assignment? You can suppress the computation of the objective function by setting the relative gap to a negative value (for example, -1). You should also simplify your function expressions as much as possible. For example, replace sub-expressions such as constantA/constantB by constantC and use put() and get() functions whenever possible. 5.7 Zone Systems 5.7.1: Can I use different zone systems in the same data bank? It is possible but not recommended. It is better to have different data banks for different zone systems. Refer to the EMME/2 User's Manual, Chapter 3, Section 3.2.2 Matrices and scenarios. 5.7.2: I need more detailed zones in some parts of my network and more aggregate data in other parts. What is the best way to modify my zone system? Refer to the EMME/2 News of March 87 - Moving matrix data between different zone systems. The basic idea is to: create 2 data banks, 1 for each zone system in each data bank, create an ensemble that will be used to define the correspondence between the 2 zone systems punch and read the matrices using this ensemble (further manipulations might be required) 5.7.3: How can I increase the number of centroids in my data bank? While it is possible to increase most of the data bank dimensions by using the macro modifydb, it is not possible to increase the number of centroids. If this is necessary, the only solution is to create a new data bank with the new dimensions. Data from the previous data bank can then be imported using Module 1.31.
5.2.1: How do I change the size of arrowheads on network plots? The size of the arrowheads can be modified in Module 1.23. The default size is 2.0. Using 0 will give links with no arrowheads. 5.2.2: How can I obtain a plot where nodes are connected by only one vector (even if two-way links exist) and with no arrowhead? Use Module 1.23 to set the size of the arrowheads to 0. In the desired modules, set the module parameter Offset to the right for links to 0. 5.2.3: Once I have a plot, how can I zoom out? Use the graphic command w followed by a number between 2 and 9. This will result in a window centered on the same node but covering an area 2 to 9 times larger. 5.2.4: In Module 2.13, how can I display the value of link or node attribute with a fixed number of decimals? You can create a new extra attribute and compute the value with the required number of decimals in Module 2.41. For example, to display the value of ul3 with 2 decimals, perform the following calculation: @att=nint(ul3*100)/100 and then, display the value of @att in Module 2.13. Starting with Release 9, you can also use the new sub-option 3 (several user defined link attributes) of option 6 (user defined data) to customize the display of link attributes, including the specification of the number of decimals. When selecting this option, you can fix the number of digits after the decimal point by specifying the selected number (y) as the second value for the Enter: Color index for layer N question following the specification of the attribute (see the EMME/2 User's Manual, Module 2.13, option 6). 5.2.5: Is it possible to display node AND link data on the same plot? No. However, it is possible to superimpose 2 plots by doing the following: set switch 18 ON to suppress the standard plot titles generate a plot with the node data outside EMME/2, transform the plot into an annotation (annotx) using GPL and the ANNOT.GPL format file (gpl -f annot.gpl -o annotx plots) generate a plot with the link data and recall annotation x (using the previous window). 5.2.6: Why is it that, when I use link color index in Module 6.12, I do not see all my links? Check that the "missing" links have a positive color index. Except in graphic worksheets, all links with a color index less than or equal to 0 are not drawn or drawn in the background color. 5.2.7: Can the information added to the plot with interactive commands in Modules 2.15, 2.43 and 6.15 be written to the plot file? Normally, this information is not sent to the plot file. However, if switch 14 is ON, all graphic output which is transmitted to the terminal is also copied to the plot file, using the terminal device definition. Note that this mode of operation is not standard and is NOT recommended for general use since it creates plot files that may be difficult to process with GPL, GPR or E2PLOTS. You can also use the Print Page option in EMTOOLW to print all the information actually on the screen. 5.2.8: How can I obtain a plot where node numbers (labels) are displayed for centroids only? Starting with Release 9, you can use the new sub-option 3 (several user defined link attributes) of option 6 (user defined data) to customize the display of link attributes, together with the specification of the subset of nodes for which the node number or label is displayed, by choosing option 3 of the following new Select question: Select: Type of node number display 1= display no node numbers 2= display all node numbers 3= display selected node numbers For more details, refer to the EMME/2 User's Manual, Module 2.13, option 6. For releases prior to Release 9, the cent2ann.awk script, available in the EMME/2, STAN Scripts section, can be used to generate an annotation file containing only the centroid numbers or labels, using the network input file (d211.in) as input. 5.3 Network and Matrix Calculators 5.3.1: How can I set color indices for different types of links? The best way is to use the network calculator (Module 2.41) to set the values in a link attribute. For example, if you want to color the connectors in green, the expression would simply be 3 (for color index 3) and, to apply the result to the connectors only, the link selection would be: ci=1 |cj=1 Note that the link attribute must first be initialized to the desired (positive) color for all other links since links with a color index of 0 are not displayed. 5.3.2: Can temporary link attributes tmplx be used in selections? No. However, it is possible to include conditions within the expression in order to "simulate" the selection. For example, if you want to change the length of the links only for those links having tmpl1 greater than 0, the expression would be: len * (tmpl1 <= 0) + expression * (tmpl1 > 0) It is also possible to replace temporary attributes by extra attributes, which can be used in the selections. 5.3.3: How can I write an expression that ensures a negative value is replaced by 0? Use the .max. operator: (expression) .max.0. 5.3.4: In the network and matrix calculators, why are the results saved when I send the report to the printer but not when I send the report to the terminal? Probably because you stopped the display of the report prematurely with a r or q. If you do not want to look at the remainder of the report, type l to skip to the end of the report. 5.4 Matrices 5.4.1: Why do some of my matrices have a "/c" flag? Probably because a transit assignment did not end normally. During a transit assignment, matrices are transformed to columnwise format in order to work destination by destination. At the end of the assignment, the matrices are retransformed to rowwise format. If the transit assignment does not end normally, the matrices will stay columnwise. Since working with columnwise matrices will increase the matrix access time drastically, you should change the storage type of columnwise matrices by using Module 3.12, option 3 (modify a matrix). 5.4.2: Matrix calculations are very slow. Is there a way to speed this up? You could do the following: Check that the matrices used in the calculations are not columnwise (see also Question 5.4.1). When running a macro with many calculations, suppress programmed pauses (~o|32). On Intel platforms, try to perform disk defragmentation. If you are using DOS or Windows 3.x, verify the size of your disk cache (Smartdrive). On UNIX platforms, try to use the data bank locally and not via a network. 5.4.3: How can I list several origin and destination matrices, excluding the zones for which the value is 0 for all matrices? Prepare an origin matrix that will be used as a constraint matrix. This matrix should contain the sum of all origin and destination matrices (destination matrices should be transposed!). For example: mo99 = mo1 + mo2 + md3'. If matrices contain negative values, the expression should be: abs(mo1) + abs(mo2) + abs(md3'). 5.5 Functions 5.5.1: How can I try different values for a constant that appears in my volume-delay function without entering the complete function each time? Put the value in a scalar matrix n and then refer to it in the function using ms(n). 5.5.2: One of the sample data files provided with the EMME/2 system contains functions. Are they general enough to be used with any data bank? No. These functions are specific to the Winnipeg data and are included solely for demonstration purposes. 5.6 Assignments 5.6.1: After an auto assignment, are the times on the connectors included in the auto time saved as a matrix? Yes. The time will depend on the functions for the connectors. 5.6.2: Is it possible to stop an auto assignment before one of the specified stopping criteria is met? Yes. Use Ctrl-C. The assignment will stop at the end of the next iteration. 5.6.3: How can I identify the non-assigned trips after an assignment? You can output the demand matrix using the time matrix as a constraint matrix with the constraint interval: (0.001,9999,exclude). This will include the trips on the diagonal (travel time of zero) and the non-connected trips (travel time equal to infinity). 5.6.4: I have a travel time matrix. How can I obtain the minimum travel time for each origin? In Module 3.21, perform a matrix computation and save the result in an origin matrix. Use the travel time matrix for the expression and as a constraint matrix (to exclude 0 values). Use the .min. operator for aggregation over destinations. 5.6.5: How can I perform an additional options auto assignment AND consider transit vehicles as additional volumes at the same time? You can compute the auto equivalent of the transit vehicles in Module 2.41 (vauteq*60/hdwy) and store the results in link and turn user data. Then, replace volad by ulN in the volume-delay functions, and pvolad by upN in the penalty functions. 5.6.6: How can I optimize the speed of my assignment? You can suppress the computation of the objective function by setting the relative gap to a negative value (for example, -1). You should also simplify your function expressions as much as possible. For example, replace sub-expressions such as constantA/constantB by constantC and use put() and get() functions whenever possible. 5.7 Zone Systems 5.7.1: Can I use different zone systems in the same data bank? It is possible but not recommended. It is better to have different data banks for different zone systems. Refer to the EMME/2 User's Manual, Chapter 3, Section 3.2.2 Matrices and scenarios. 5.7.2: I need more detailed zones in some parts of my network and more aggregate data in other parts. What is the best way to modify my zone system? Refer to the EMME/2 News of March 87 - Moving matrix data between different zone systems. The basic idea is to: create 2 data banks, 1 for each zone system in each data bank, create an ensemble that will be used to define the correspondence between the 2 zone systems punch and read the matrices using this ensemble (further manipulations might be required) 5.7.3: How can I increase the number of centroids in my data bank? While it is possible to increase most of the data bank dimensions by using the macro modifydb, it is not possible to increase the number of centroids. If this is necessary, the only solution is to create a new data bank with the new dimensions. Data from the previous data bank can then be imported using Module 1.31.
The size of the arrowheads can be modified in Module 1.23. The default size is 2.0. Using 0 will give links with no arrowheads.
5.2.2: How can I obtain a plot where nodes are connected by only one vector (even if two-way links exist) and with no arrowhead? Use Module 1.23 to set the size of the arrowheads to 0. In the desired modules, set the module parameter Offset to the right for links to 0. 5.2.3: Once I have a plot, how can I zoom out? Use the graphic command w followed by a number between 2 and 9. This will result in a window centered on the same node but covering an area 2 to 9 times larger. 5.2.4: In Module 2.13, how can I display the value of link or node attribute with a fixed number of decimals? You can create a new extra attribute and compute the value with the required number of decimals in Module 2.41. For example, to display the value of ul3 with 2 decimals, perform the following calculation: @att=nint(ul3*100)/100 and then, display the value of @att in Module 2.13. Starting with Release 9, you can also use the new sub-option 3 (several user defined link attributes) of option 6 (user defined data) to customize the display of link attributes, including the specification of the number of decimals. When selecting this option, you can fix the number of digits after the decimal point by specifying the selected number (y) as the second value for the Enter: Color index for layer N question following the specification of the attribute (see the EMME/2 User's Manual, Module 2.13, option 6). 5.2.5: Is it possible to display node AND link data on the same plot? No. However, it is possible to superimpose 2 plots by doing the following: set switch 18 ON to suppress the standard plot titles generate a plot with the node data outside EMME/2, transform the plot into an annotation (annotx) using GPL and the ANNOT.GPL format file (gpl -f annot.gpl -o annotx plots) generate a plot with the link data and recall annotation x (using the previous window). 5.2.6: Why is it that, when I use link color index in Module 6.12, I do not see all my links? Check that the "missing" links have a positive color index. Except in graphic worksheets, all links with a color index less than or equal to 0 are not drawn or drawn in the background color. 5.2.7: Can the information added to the plot with interactive commands in Modules 2.15, 2.43 and 6.15 be written to the plot file? Normally, this information is not sent to the plot file. However, if switch 14 is ON, all graphic output which is transmitted to the terminal is also copied to the plot file, using the terminal device definition. Note that this mode of operation is not standard and is NOT recommended for general use since it creates plot files that may be difficult to process with GPL, GPR or E2PLOTS. You can also use the Print Page option in EMTOOLW to print all the information actually on the screen. 5.2.8: How can I obtain a plot where node numbers (labels) are displayed for centroids only? Starting with Release 9, you can use the new sub-option 3 (several user defined link attributes) of option 6 (user defined data) to customize the display of link attributes, together with the specification of the subset of nodes for which the node number or label is displayed, by choosing option 3 of the following new Select question: Select: Type of node number display 1= display no node numbers 2= display all node numbers 3= display selected node numbers For more details, refer to the EMME/2 User's Manual, Module 2.13, option 6. For releases prior to Release 9, the cent2ann.awk script, available in the EMME/2, STAN Scripts section, can be used to generate an annotation file containing only the centroid numbers or labels, using the network input file (d211.in) as input. 5.3 Network and Matrix Calculators 5.3.1: How can I set color indices for different types of links? The best way is to use the network calculator (Module 2.41) to set the values in a link attribute. For example, if you want to color the connectors in green, the expression would simply be 3 (for color index 3) and, to apply the result to the connectors only, the link selection would be: ci=1 |cj=1 Note that the link attribute must first be initialized to the desired (positive) color for all other links since links with a color index of 0 are not displayed. 5.3.2: Can temporary link attributes tmplx be used in selections? No. However, it is possible to include conditions within the expression in order to "simulate" the selection. For example, if you want to change the length of the links only for those links having tmpl1 greater than 0, the expression would be: len * (tmpl1 <= 0) + expression * (tmpl1 > 0) It is also possible to replace temporary attributes by extra attributes, which can be used in the selections. 5.3.3: How can I write an expression that ensures a negative value is replaced by 0? Use the .max. operator: (expression) .max.0. 5.3.4: In the network and matrix calculators, why are the results saved when I send the report to the printer but not when I send the report to the terminal? Probably because you stopped the display of the report prematurely with a r or q. If you do not want to look at the remainder of the report, type l to skip to the end of the report. 5.4 Matrices 5.4.1: Why do some of my matrices have a "/c" flag? Probably because a transit assignment did not end normally. During a transit assignment, matrices are transformed to columnwise format in order to work destination by destination. At the end of the assignment, the matrices are retransformed to rowwise format. If the transit assignment does not end normally, the matrices will stay columnwise. Since working with columnwise matrices will increase the matrix access time drastically, you should change the storage type of columnwise matrices by using Module 3.12, option 3 (modify a matrix). 5.4.2: Matrix calculations are very slow. Is there a way to speed this up? You could do the following: Check that the matrices used in the calculations are not columnwise (see also Question 5.4.1). When running a macro with many calculations, suppress programmed pauses (~o|32). On Intel platforms, try to perform disk defragmentation. If you are using DOS or Windows 3.x, verify the size of your disk cache (Smartdrive). On UNIX platforms, try to use the data bank locally and not via a network. 5.4.3: How can I list several origin and destination matrices, excluding the zones for which the value is 0 for all matrices? Prepare an origin matrix that will be used as a constraint matrix. This matrix should contain the sum of all origin and destination matrices (destination matrices should be transposed!). For example: mo99 = mo1 + mo2 + md3'. If matrices contain negative values, the expression should be: abs(mo1) + abs(mo2) + abs(md3'). 5.5 Functions 5.5.1: How can I try different values for a constant that appears in my volume-delay function without entering the complete function each time? Put the value in a scalar matrix n and then refer to it in the function using ms(n). 5.5.2: One of the sample data files provided with the EMME/2 system contains functions. Are they general enough to be used with any data bank? No. These functions are specific to the Winnipeg data and are included solely for demonstration purposes. 5.6 Assignments 5.6.1: After an auto assignment, are the times on the connectors included in the auto time saved as a matrix? Yes. The time will depend on the functions for the connectors. 5.6.2: Is it possible to stop an auto assignment before one of the specified stopping criteria is met? Yes. Use Ctrl-C. The assignment will stop at the end of the next iteration. 5.6.3: How can I identify the non-assigned trips after an assignment? You can output the demand matrix using the time matrix as a constraint matrix with the constraint interval: (0.001,9999,exclude). This will include the trips on the diagonal (travel time of zero) and the non-connected trips (travel time equal to infinity). 5.6.4: I have a travel time matrix. How can I obtain the minimum travel time for each origin? In Module 3.21, perform a matrix computation and save the result in an origin matrix. Use the travel time matrix for the expression and as a constraint matrix (to exclude 0 values). Use the .min. operator for aggregation over destinations. 5.6.5: How can I perform an additional options auto assignment AND consider transit vehicles as additional volumes at the same time? You can compute the auto equivalent of the transit vehicles in Module 2.41 (vauteq*60/hdwy) and store the results in link and turn user data. Then, replace volad by ulN in the volume-delay functions, and pvolad by upN in the penalty functions. 5.6.6: How can I optimize the speed of my assignment? You can suppress the computation of the objective function by setting the relative gap to a negative value (for example, -1). You should also simplify your function expressions as much as possible. For example, replace sub-expressions such as constantA/constantB by constantC and use put() and get() functions whenever possible. 5.7 Zone Systems 5.7.1: Can I use different zone systems in the same data bank? It is possible but not recommended. It is better to have different data banks for different zone systems. Refer to the EMME/2 User's Manual, Chapter 3, Section 3.2.2 Matrices and scenarios. 5.7.2: I need more detailed zones in some parts of my network and more aggregate data in other parts. What is the best way to modify my zone system? Refer to the EMME/2 News of March 87 - Moving matrix data between different zone systems. The basic idea is to: create 2 data banks, 1 for each zone system in each data bank, create an ensemble that will be used to define the correspondence between the 2 zone systems punch and read the matrices using this ensemble (further manipulations might be required) 5.7.3: How can I increase the number of centroids in my data bank? While it is possible to increase most of the data bank dimensions by using the macro modifydb, it is not possible to increase the number of centroids. If this is necessary, the only solution is to create a new data bank with the new dimensions. Data from the previous data bank can then be imported using Module 1.31.
Use Module 1.23 to set the size of the arrowheads to 0. In the desired modules, set the module parameter Offset to the right for links to 0.
5.2.3: Once I have a plot, how can I zoom out? Use the graphic command w followed by a number between 2 and 9. This will result in a window centered on the same node but covering an area 2 to 9 times larger. 5.2.4: In Module 2.13, how can I display the value of link or node attribute with a fixed number of decimals? You can create a new extra attribute and compute the value with the required number of decimals in Module 2.41. For example, to display the value of ul3 with 2 decimals, perform the following calculation: @att=nint(ul3*100)/100 and then, display the value of @att in Module 2.13. Starting with Release 9, you can also use the new sub-option 3 (several user defined link attributes) of option 6 (user defined data) to customize the display of link attributes, including the specification of the number of decimals. When selecting this option, you can fix the number of digits after the decimal point by specifying the selected number (y) as the second value for the Enter: Color index for layer N question following the specification of the attribute (see the EMME/2 User's Manual, Module 2.13, option 6). 5.2.5: Is it possible to display node AND link data on the same plot? No. However, it is possible to superimpose 2 plots by doing the following: set switch 18 ON to suppress the standard plot titles generate a plot with the node data outside EMME/2, transform the plot into an annotation (annotx) using GPL and the ANNOT.GPL format file (gpl -f annot.gpl -o annotx plots) generate a plot with the link data and recall annotation x (using the previous window). 5.2.6: Why is it that, when I use link color index in Module 6.12, I do not see all my links? Check that the "missing" links have a positive color index. Except in graphic worksheets, all links with a color index less than or equal to 0 are not drawn or drawn in the background color. 5.2.7: Can the information added to the plot with interactive commands in Modules 2.15, 2.43 and 6.15 be written to the plot file? Normally, this information is not sent to the plot file. However, if switch 14 is ON, all graphic output which is transmitted to the terminal is also copied to the plot file, using the terminal device definition. Note that this mode of operation is not standard and is NOT recommended for general use since it creates plot files that may be difficult to process with GPL, GPR or E2PLOTS. You can also use the Print Page option in EMTOOLW to print all the information actually on the screen. 5.2.8: How can I obtain a plot where node numbers (labels) are displayed for centroids only? Starting with Release 9, you can use the new sub-option 3 (several user defined link attributes) of option 6 (user defined data) to customize the display of link attributes, together with the specification of the subset of nodes for which the node number or label is displayed, by choosing option 3 of the following new Select question: Select: Type of node number display 1= display no node numbers 2= display all node numbers 3= display selected node numbers For more details, refer to the EMME/2 User's Manual, Module 2.13, option 6. For releases prior to Release 9, the cent2ann.awk script, available in the EMME/2, STAN Scripts section, can be used to generate an annotation file containing only the centroid numbers or labels, using the network input file (d211.in) as input. 5.3 Network and Matrix Calculators 5.3.1: How can I set color indices for different types of links? The best way is to use the network calculator (Module 2.41) to set the values in a link attribute. For example, if you want to color the connectors in green, the expression would simply be 3 (for color index 3) and, to apply the result to the connectors only, the link selection would be: ci=1 |cj=1 Note that the link attribute must first be initialized to the desired (positive) color for all other links since links with a color index of 0 are not displayed. 5.3.2: Can temporary link attributes tmplx be used in selections? No. However, it is possible to include conditions within the expression in order to "simulate" the selection. For example, if you want to change the length of the links only for those links having tmpl1 greater than 0, the expression would be: len * (tmpl1 <= 0) + expression * (tmpl1 > 0) It is also possible to replace temporary attributes by extra attributes, which can be used in the selections. 5.3.3: How can I write an expression that ensures a negative value is replaced by 0? Use the .max. operator: (expression) .max.0. 5.3.4: In the network and matrix calculators, why are the results saved when I send the report to the printer but not when I send the report to the terminal? Probably because you stopped the display of the report prematurely with a r or q. If you do not want to look at the remainder of the report, type l to skip to the end of the report. 5.4 Matrices 5.4.1: Why do some of my matrices have a "/c" flag? Probably because a transit assignment did not end normally. During a transit assignment, matrices are transformed to columnwise format in order to work destination by destination. At the end of the assignment, the matrices are retransformed to rowwise format. If the transit assignment does not end normally, the matrices will stay columnwise. Since working with columnwise matrices will increase the matrix access time drastically, you should change the storage type of columnwise matrices by using Module 3.12, option 3 (modify a matrix). 5.4.2: Matrix calculations are very slow. Is there a way to speed this up? You could do the following: Check that the matrices used in the calculations are not columnwise (see also Question 5.4.1). When running a macro with many calculations, suppress programmed pauses (~o|32). On Intel platforms, try to perform disk defragmentation. If you are using DOS or Windows 3.x, verify the size of your disk cache (Smartdrive). On UNIX platforms, try to use the data bank locally and not via a network. 5.4.3: How can I list several origin and destination matrices, excluding the zones for which the value is 0 for all matrices? Prepare an origin matrix that will be used as a constraint matrix. This matrix should contain the sum of all origin and destination matrices (destination matrices should be transposed!). For example: mo99 = mo1 + mo2 + md3'. If matrices contain negative values, the expression should be: abs(mo1) + abs(mo2) + abs(md3'). 5.5 Functions 5.5.1: How can I try different values for a constant that appears in my volume-delay function without entering the complete function each time? Put the value in a scalar matrix n and then refer to it in the function using ms(n). 5.5.2: One of the sample data files provided with the EMME/2 system contains functions. Are they general enough to be used with any data bank? No. These functions are specific to the Winnipeg data and are included solely for demonstration purposes. 5.6 Assignments 5.6.1: After an auto assignment, are the times on the connectors included in the auto time saved as a matrix? Yes. The time will depend on the functions for the connectors. 5.6.2: Is it possible to stop an auto assignment before one of the specified stopping criteria is met? Yes. Use Ctrl-C. The assignment will stop at the end of the next iteration. 5.6.3: How can I identify the non-assigned trips after an assignment? You can output the demand matrix using the time matrix as a constraint matrix with the constraint interval: (0.001,9999,exclude). This will include the trips on the diagonal (travel time of zero) and the non-connected trips (travel time equal to infinity). 5.6.4: I have a travel time matrix. How can I obtain the minimum travel time for each origin? In Module 3.21, perform a matrix computation and save the result in an origin matrix. Use the travel time matrix for the expression and as a constraint matrix (to exclude 0 values). Use the .min. operator for aggregation over destinations. 5.6.5: How can I perform an additional options auto assignment AND consider transit vehicles as additional volumes at the same time? You can compute the auto equivalent of the transit vehicles in Module 2.41 (vauteq*60/hdwy) and store the results in link and turn user data. Then, replace volad by ulN in the volume-delay functions, and pvolad by upN in the penalty functions. 5.6.6: How can I optimize the speed of my assignment? You can suppress the computation of the objective function by setting the relative gap to a negative value (for example, -1). You should also simplify your function expressions as much as possible. For example, replace sub-expressions such as constantA/constantB by constantC and use put() and get() functions whenever possible. 5.7 Zone Systems 5.7.1: Can I use different zone systems in the same data bank? It is possible but not recommended. It is better to have different data banks for different zone systems. Refer to the EMME/2 User's Manual, Chapter 3, Section 3.2.2 Matrices and scenarios. 5.7.2: I need more detailed zones in some parts of my network and more aggregate data in other parts. What is the best way to modify my zone system? Refer to the EMME/2 News of March 87 - Moving matrix data between different zone systems. The basic idea is to: create 2 data banks, 1 for each zone system in each data bank, create an ensemble that will be used to define the correspondence between the 2 zone systems punch and read the matrices using this ensemble (further manipulations might be required) 5.7.3: How can I increase the number of centroids in my data bank? While it is possible to increase most of the data bank dimensions by using the macro modifydb, it is not possible to increase the number of centroids. If this is necessary, the only solution is to create a new data bank with the new dimensions. Data from the previous data bank can then be imported using Module 1.31.
Use the graphic command w followed by a number between 2 and 9. This will result in a window centered on the same node but covering an area 2 to 9 times larger.
5.2.4: In Module 2.13, how can I display the value of link or node attribute with a fixed number of decimals? You can create a new extra attribute and compute the value with the required number of decimals in Module 2.41. For example, to display the value of ul3 with 2 decimals, perform the following calculation: @att=nint(ul3*100)/100 and then, display the value of @att in Module 2.13. Starting with Release 9, you can also use the new sub-option 3 (several user defined link attributes) of option 6 (user defined data) to customize the display of link attributes, including the specification of the number of decimals. When selecting this option, you can fix the number of digits after the decimal point by specifying the selected number (y) as the second value for the Enter: Color index for layer N question following the specification of the attribute (see the EMME/2 User's Manual, Module 2.13, option 6). 5.2.5: Is it possible to display node AND link data on the same plot? No. However, it is possible to superimpose 2 plots by doing the following: set switch 18 ON to suppress the standard plot titles generate a plot with the node data outside EMME/2, transform the plot into an annotation (annotx) using GPL and the ANNOT.GPL format file (gpl -f annot.gpl -o annotx plots) generate a plot with the link data and recall annotation x (using the previous window). 5.2.6: Why is it that, when I use link color index in Module 6.12, I do not see all my links? Check that the "missing" links have a positive color index. Except in graphic worksheets, all links with a color index less than or equal to 0 are not drawn or drawn in the background color. 5.2.7: Can the information added to the plot with interactive commands in Modules 2.15, 2.43 and 6.15 be written to the plot file? Normally, this information is not sent to the plot file. However, if switch 14 is ON, all graphic output which is transmitted to the terminal is also copied to the plot file, using the terminal device definition. Note that this mode of operation is not standard and is NOT recommended for general use since it creates plot files that may be difficult to process with GPL, GPR or E2PLOTS. You can also use the Print Page option in EMTOOLW to print all the information actually on the screen. 5.2.8: How can I obtain a plot where node numbers (labels) are displayed for centroids only? Starting with Release 9, you can use the new sub-option 3 (several user defined link attributes) of option 6 (user defined data) to customize the display of link attributes, together with the specification of the subset of nodes for which the node number or label is displayed, by choosing option 3 of the following new Select question: Select: Type of node number display 1= display no node numbers 2= display all node numbers 3= display selected node numbers For more details, refer to the EMME/2 User's Manual, Module 2.13, option 6. For releases prior to Release 9, the cent2ann.awk script, available in the EMME/2, STAN Scripts section, can be used to generate an annotation file containing only the centroid numbers or labels, using the network input file (d211.in) as input. 5.3 Network and Matrix Calculators 5.3.1: How can I set color indices for different types of links? The best way is to use the network calculator (Module 2.41) to set the values in a link attribute. For example, if you want to color the connectors in green, the expression would simply be 3 (for color index 3) and, to apply the result to the connectors only, the link selection would be: ci=1 |cj=1 Note that the link attribute must first be initialized to the desired (positive) color for all other links since links with a color index of 0 are not displayed. 5.3.2: Can temporary link attributes tmplx be used in selections? No. However, it is possible to include conditions within the expression in order to "simulate" the selection. For example, if you want to change the length of the links only for those links having tmpl1 greater than 0, the expression would be: len * (tmpl1 <= 0) + expression * (tmpl1 > 0) It is also possible to replace temporary attributes by extra attributes, which can be used in the selections. 5.3.3: How can I write an expression that ensures a negative value is replaced by 0? Use the .max. operator: (expression) .max.0. 5.3.4: In the network and matrix calculators, why are the results saved when I send the report to the printer but not when I send the report to the terminal? Probably because you stopped the display of the report prematurely with a r or q. If you do not want to look at the remainder of the report, type l to skip to the end of the report. 5.4 Matrices 5.4.1: Why do some of my matrices have a "/c" flag? Probably because a transit assignment did not end normally. During a transit assignment, matrices are transformed to columnwise format in order to work destination by destination. At the end of the assignment, the matrices are retransformed to rowwise format. If the transit assignment does not end normally, the matrices will stay columnwise. Since working with columnwise matrices will increase the matrix access time drastically, you should change the storage type of columnwise matrices by using Module 3.12, option 3 (modify a matrix). 5.4.2: Matrix calculations are very slow. Is there a way to speed this up? You could do the following: Check that the matrices used in the calculations are not columnwise (see also Question 5.4.1). When running a macro with many calculations, suppress programmed pauses (~o|32). On Intel platforms, try to perform disk defragmentation. If you are using DOS or Windows 3.x, verify the size of your disk cache (Smartdrive). On UNIX platforms, try to use the data bank locally and not via a network. 5.4.3: How can I list several origin and destination matrices, excluding the zones for which the value is 0 for all matrices? Prepare an origin matrix that will be used as a constraint matrix. This matrix should contain the sum of all origin and destination matrices (destination matrices should be transposed!). For example: mo99 = mo1 + mo2 + md3'. If matrices contain negative values, the expression should be: abs(mo1) + abs(mo2) + abs(md3'). 5.5 Functions 5.5.1: How can I try different values for a constant that appears in my volume-delay function without entering the complete function each time? Put the value in a scalar matrix n and then refer to it in the function using ms(n). 5.5.2: One of the sample data files provided with the EMME/2 system contains functions. Are they general enough to be used with any data bank? No. These functions are specific to the Winnipeg data and are included solely for demonstration purposes. 5.6 Assignments 5.6.1: After an auto assignment, are the times on the connectors included in the auto time saved as a matrix? Yes. The time will depend on the functions for the connectors. 5.6.2: Is it possible to stop an auto assignment before one of the specified stopping criteria is met? Yes. Use Ctrl-C. The assignment will stop at the end of the next iteration. 5.6.3: How can I identify the non-assigned trips after an assignment? You can output the demand matrix using the time matrix as a constraint matrix with the constraint interval: (0.001,9999,exclude). This will include the trips on the diagonal (travel time of zero) and the non-connected trips (travel time equal to infinity). 5.6.4: I have a travel time matrix. How can I obtain the minimum travel time for each origin? In Module 3.21, perform a matrix computation and save the result in an origin matrix. Use the travel time matrix for the expression and as a constraint matrix (to exclude 0 values). Use the .min. operator for aggregation over destinations. 5.6.5: How can I perform an additional options auto assignment AND consider transit vehicles as additional volumes at the same time? You can compute the auto equivalent of the transit vehicles in Module 2.41 (vauteq*60/hdwy) and store the results in link and turn user data. Then, replace volad by ulN in the volume-delay functions, and pvolad by upN in the penalty functions. 5.6.6: How can I optimize the speed of my assignment? You can suppress the computation of the objective function by setting the relative gap to a negative value (for example, -1). You should also simplify your function expressions as much as possible. For example, replace sub-expressions such as constantA/constantB by constantC and use put() and get() functions whenever possible. 5.7 Zone Systems 5.7.1: Can I use different zone systems in the same data bank? It is possible but not recommended. It is better to have different data banks for different zone systems. Refer to the EMME/2 User's Manual, Chapter 3, Section 3.2.2 Matrices and scenarios. 5.7.2: I need more detailed zones in some parts of my network and more aggregate data in other parts. What is the best way to modify my zone system? Refer to the EMME/2 News of March 87 - Moving matrix data between different zone systems. The basic idea is to: create 2 data banks, 1 for each zone system in each data bank, create an ensemble that will be used to define the correspondence between the 2 zone systems punch and read the matrices using this ensemble (further manipulations might be required) 5.7.3: How can I increase the number of centroids in my data bank? While it is possible to increase most of the data bank dimensions by using the macro modifydb, it is not possible to increase the number of centroids. If this is necessary, the only solution is to create a new data bank with the new dimensions. Data from the previous data bank can then be imported using Module 1.31.
You can create a new extra attribute and compute the value with the required number of decimals in Module 2.41. For example, to display the value of ul3 with 2 decimals, perform the following calculation:
and then, display the value of @att in Module 2.13.
Starting with Release 9, you can also use the new sub-option 3 (several user defined link attributes) of option 6 (user defined data) to customize the display of link attributes, including the specification of the number of decimals. When selecting this option, you can fix the number of digits after the decimal point by specifying the selected number (y) as the second value for the Enter: Color index for layer N question following the specification of the attribute (see the EMME/2 User's Manual, Module 2.13, option 6).
5.2.5: Is it possible to display node AND link data on the same plot? No. However, it is possible to superimpose 2 plots by doing the following: set switch 18 ON to suppress the standard plot titles generate a plot with the node data outside EMME/2, transform the plot into an annotation (annotx) using GPL and the ANNOT.GPL format file (gpl -f annot.gpl -o annotx plots) generate a plot with the link data and recall annotation x (using the previous window). 5.2.6: Why is it that, when I use link color index in Module 6.12, I do not see all my links? Check that the "missing" links have a positive color index. Except in graphic worksheets, all links with a color index less than or equal to 0 are not drawn or drawn in the background color. 5.2.7: Can the information added to the plot with interactive commands in Modules 2.15, 2.43 and 6.15 be written to the plot file? Normally, this information is not sent to the plot file. However, if switch 14 is ON, all graphic output which is transmitted to the terminal is also copied to the plot file, using the terminal device definition. Note that this mode of operation is not standard and is NOT recommended for general use since it creates plot files that may be difficult to process with GPL, GPR or E2PLOTS. You can also use the Print Page option in EMTOOLW to print all the information actually on the screen. 5.2.8: How can I obtain a plot where node numbers (labels) are displayed for centroids only? Starting with Release 9, you can use the new sub-option 3 (several user defined link attributes) of option 6 (user defined data) to customize the display of link attributes, together with the specification of the subset of nodes for which the node number or label is displayed, by choosing option 3 of the following new Select question: Select: Type of node number display 1= display no node numbers 2= display all node numbers 3= display selected node numbers For more details, refer to the EMME/2 User's Manual, Module 2.13, option 6. For releases prior to Release 9, the cent2ann.awk script, available in the EMME/2, STAN Scripts section, can be used to generate an annotation file containing only the centroid numbers or labels, using the network input file (d211.in) as input. 5.3 Network and Matrix Calculators 5.3.1: How can I set color indices for different types of links? The best way is to use the network calculator (Module 2.41) to set the values in a link attribute. For example, if you want to color the connectors in green, the expression would simply be 3 (for color index 3) and, to apply the result to the connectors only, the link selection would be: ci=1 |cj=1 Note that the link attribute must first be initialized to the desired (positive) color for all other links since links with a color index of 0 are not displayed. 5.3.2: Can temporary link attributes tmplx be used in selections? No. However, it is possible to include conditions within the expression in order to "simulate" the selection. For example, if you want to change the length of the links only for those links having tmpl1 greater than 0, the expression would be: len * (tmpl1 <= 0) + expression * (tmpl1 > 0) It is also possible to replace temporary attributes by extra attributes, which can be used in the selections. 5.3.3: How can I write an expression that ensures a negative value is replaced by 0? Use the .max. operator: (expression) .max.0. 5.3.4: In the network and matrix calculators, why are the results saved when I send the report to the printer but not when I send the report to the terminal? Probably because you stopped the display of the report prematurely with a r or q. If you do not want to look at the remainder of the report, type l to skip to the end of the report. 5.4 Matrices 5.4.1: Why do some of my matrices have a "/c" flag? Probably because a transit assignment did not end normally. During a transit assignment, matrices are transformed to columnwise format in order to work destination by destination. At the end of the assignment, the matrices are retransformed to rowwise format. If the transit assignment does not end normally, the matrices will stay columnwise. Since working with columnwise matrices will increase the matrix access time drastically, you should change the storage type of columnwise matrices by using Module 3.12, option 3 (modify a matrix). 5.4.2: Matrix calculations are very slow. Is there a way to speed this up? You could do the following: Check that the matrices used in the calculations are not columnwise (see also Question 5.4.1). When running a macro with many calculations, suppress programmed pauses (~o|32). On Intel platforms, try to perform disk defragmentation. If you are using DOS or Windows 3.x, verify the size of your disk cache (Smartdrive). On UNIX platforms, try to use the data bank locally and not via a network. 5.4.3: How can I list several origin and destination matrices, excluding the zones for which the value is 0 for all matrices? Prepare an origin matrix that will be used as a constraint matrix. This matrix should contain the sum of all origin and destination matrices (destination matrices should be transposed!). For example: mo99 = mo1 + mo2 + md3'. If matrices contain negative values, the expression should be: abs(mo1) + abs(mo2) + abs(md3'). 5.5 Functions 5.5.1: How can I try different values for a constant that appears in my volume-delay function without entering the complete function each time? Put the value in a scalar matrix n and then refer to it in the function using ms(n). 5.5.2: One of the sample data files provided with the EMME/2 system contains functions. Are they general enough to be used with any data bank? No. These functions are specific to the Winnipeg data and are included solely for demonstration purposes. 5.6 Assignments 5.6.1: After an auto assignment, are the times on the connectors included in the auto time saved as a matrix? Yes. The time will depend on the functions for the connectors. 5.6.2: Is it possible to stop an auto assignment before one of the specified stopping criteria is met? Yes. Use Ctrl-C. The assignment will stop at the end of the next iteration. 5.6.3: How can I identify the non-assigned trips after an assignment? You can output the demand matrix using the time matrix as a constraint matrix with the constraint interval: (0.001,9999,exclude). This will include the trips on the diagonal (travel time of zero) and the non-connected trips (travel time equal to infinity). 5.6.4: I have a travel time matrix. How can I obtain the minimum travel time for each origin? In Module 3.21, perform a matrix computation and save the result in an origin matrix. Use the travel time matrix for the expression and as a constraint matrix (to exclude 0 values). Use the .min. operator for aggregation over destinations. 5.6.5: How can I perform an additional options auto assignment AND consider transit vehicles as additional volumes at the same time? You can compute the auto equivalent of the transit vehicles in Module 2.41 (vauteq*60/hdwy) and store the results in link and turn user data. Then, replace volad by ulN in the volume-delay functions, and pvolad by upN in the penalty functions. 5.6.6: How can I optimize the speed of my assignment? You can suppress the computation of the objective function by setting the relative gap to a negative value (for example, -1). You should also simplify your function expressions as much as possible. For example, replace sub-expressions such as constantA/constantB by constantC and use put() and get() functions whenever possible. 5.7 Zone Systems 5.7.1: Can I use different zone systems in the same data bank? It is possible but not recommended. It is better to have different data banks for different zone systems. Refer to the EMME/2 User's Manual, Chapter 3, Section 3.2.2 Matrices and scenarios. 5.7.2: I need more detailed zones in some parts of my network and more aggregate data in other parts. What is the best way to modify my zone system? Refer to the EMME/2 News of March 87 - Moving matrix data between different zone systems. The basic idea is to: create 2 data banks, 1 for each zone system in each data bank, create an ensemble that will be used to define the correspondence between the 2 zone systems punch and read the matrices using this ensemble (further manipulations might be required) 5.7.3: How can I increase the number of centroids in my data bank? While it is possible to increase most of the data bank dimensions by using the macro modifydb, it is not possible to increase the number of centroids. If this is necessary, the only solution is to create a new data bank with the new dimensions. Data from the previous data bank can then be imported using Module 1.31.
No. However, it is possible to superimpose 2 plots by doing the following:
5.2.6: Why is it that, when I use link color index in Module 6.12, I do not see all my links? Check that the "missing" links have a positive color index. Except in graphic worksheets, all links with a color index less than or equal to 0 are not drawn or drawn in the background color. 5.2.7: Can the information added to the plot with interactive commands in Modules 2.15, 2.43 and 6.15 be written to the plot file? Normally, this information is not sent to the plot file. However, if switch 14 is ON, all graphic output which is transmitted to the terminal is also copied to the plot file, using the terminal device definition. Note that this mode of operation is not standard and is NOT recommended for general use since it creates plot files that may be difficult to process with GPL, GPR or E2PLOTS. You can also use the Print Page option in EMTOOLW to print all the information actually on the screen. 5.2.8: How can I obtain a plot where node numbers (labels) are displayed for centroids only? Starting with Release 9, you can use the new sub-option 3 (several user defined link attributes) of option 6 (user defined data) to customize the display of link attributes, together with the specification of the subset of nodes for which the node number or label is displayed, by choosing option 3 of the following new Select question: Select: Type of node number display 1= display no node numbers 2= display all node numbers 3= display selected node numbers For more details, refer to the EMME/2 User's Manual, Module 2.13, option 6. For releases prior to Release 9, the cent2ann.awk script, available in the EMME/2, STAN Scripts section, can be used to generate an annotation file containing only the centroid numbers or labels, using the network input file (d211.in) as input. 5.3 Network and Matrix Calculators 5.3.1: How can I set color indices for different types of links? The best way is to use the network calculator (Module 2.41) to set the values in a link attribute. For example, if you want to color the connectors in green, the expression would simply be 3 (for color index 3) and, to apply the result to the connectors only, the link selection would be: ci=1 |cj=1 Note that the link attribute must first be initialized to the desired (positive) color for all other links since links with a color index of 0 are not displayed. 5.3.2: Can temporary link attributes tmplx be used in selections? No. However, it is possible to include conditions within the expression in order to "simulate" the selection. For example, if you want to change the length of the links only for those links having tmpl1 greater than 0, the expression would be: len * (tmpl1 <= 0) + expression * (tmpl1 > 0) It is also possible to replace temporary attributes by extra attributes, which can be used in the selections. 5.3.3: How can I write an expression that ensures a negative value is replaced by 0? Use the .max. operator: (expression) .max.0. 5.3.4: In the network and matrix calculators, why are the results saved when I send the report to the printer but not when I send the report to the terminal? Probably because you stopped the display of the report prematurely with a r or q. If you do not want to look at the remainder of the report, type l to skip to the end of the report. 5.4 Matrices 5.4.1: Why do some of my matrices have a "/c" flag? Probably because a transit assignment did not end normally. During a transit assignment, matrices are transformed to columnwise format in order to work destination by destination. At the end of the assignment, the matrices are retransformed to rowwise format. If the transit assignment does not end normally, the matrices will stay columnwise. Since working with columnwise matrices will increase the matrix access time drastically, you should change the storage type of columnwise matrices by using Module 3.12, option 3 (modify a matrix). 5.4.2: Matrix calculations are very slow. Is there a way to speed this up? You could do the following: Check that the matrices used in the calculations are not columnwise (see also Question 5.4.1). When running a macro with many calculations, suppress programmed pauses (~o|32). On Intel platforms, try to perform disk defragmentation. If you are using DOS or Windows 3.x, verify the size of your disk cache (Smartdrive). On UNIX platforms, try to use the data bank locally and not via a network. 5.4.3: How can I list several origin and destination matrices, excluding the zones for which the value is 0 for all matrices? Prepare an origin matrix that will be used as a constraint matrix. This matrix should contain the sum of all origin and destination matrices (destination matrices should be transposed!). For example: mo99 = mo1 + mo2 + md3'. If matrices contain negative values, the expression should be: abs(mo1) + abs(mo2) + abs(md3'). 5.5 Functions 5.5.1: How can I try different values for a constant that appears in my volume-delay function without entering the complete function each time? Put the value in a scalar matrix n and then refer to it in the function using ms(n). 5.5.2: One of the sample data files provided with the EMME/2 system contains functions. Are they general enough to be used with any data bank? No. These functions are specific to the Winnipeg data and are included solely for demonstration purposes. 5.6 Assignments 5.6.1: After an auto assignment, are the times on the connectors included in the auto time saved as a matrix? Yes. The time will depend on the functions for the connectors. 5.6.2: Is it possible to stop an auto assignment before one of the specified stopping criteria is met? Yes. Use Ctrl-C. The assignment will stop at the end of the next iteration. 5.6.3: How can I identify the non-assigned trips after an assignment? You can output the demand matrix using the time matrix as a constraint matrix with the constraint interval: (0.001,9999,exclude). This will include the trips on the diagonal (travel time of zero) and the non-connected trips (travel time equal to infinity). 5.6.4: I have a travel time matrix. How can I obtain the minimum travel time for each origin? In Module 3.21, perform a matrix computation and save the result in an origin matrix. Use the travel time matrix for the expression and as a constraint matrix (to exclude 0 values). Use the .min. operator for aggregation over destinations. 5.6.5: How can I perform an additional options auto assignment AND consider transit vehicles as additional volumes at the same time? You can compute the auto equivalent of the transit vehicles in Module 2.41 (vauteq*60/hdwy) and store the results in link and turn user data. Then, replace volad by ulN in the volume-delay functions, and pvolad by upN in the penalty functions. 5.6.6: How can I optimize the speed of my assignment? You can suppress the computation of the objective function by setting the relative gap to a negative value (for example, -1). You should also simplify your function expressions as much as possible. For example, replace sub-expressions such as constantA/constantB by constantC and use put() and get() functions whenever possible. 5.7 Zone Systems 5.7.1: Can I use different zone systems in the same data bank? It is possible but not recommended. It is better to have different data banks for different zone systems. Refer to the EMME/2 User's Manual, Chapter 3, Section 3.2.2 Matrices and scenarios. 5.7.2: I need more detailed zones in some parts of my network and more aggregate data in other parts. What is the best way to modify my zone system? Refer to the EMME/2 News of March 87 - Moving matrix data between different zone systems. The basic idea is to: create 2 data banks, 1 for each zone system in each data bank, create an ensemble that will be used to define the correspondence between the 2 zone systems punch and read the matrices using this ensemble (further manipulations might be required) 5.7.3: How can I increase the number of centroids in my data bank? While it is possible to increase most of the data bank dimensions by using the macro modifydb, it is not possible to increase the number of centroids. If this is necessary, the only solution is to create a new data bank with the new dimensions. Data from the previous data bank can then be imported using Module 1.31.
Check that the "missing" links have a positive color index. Except in graphic worksheets, all links with a color index less than or equal to 0 are not drawn or drawn in the background color.
5.2.7: Can the information added to the plot with interactive commands in Modules 2.15, 2.43 and 6.15 be written to the plot file? Normally, this information is not sent to the plot file. However, if switch 14 is ON, all graphic output which is transmitted to the terminal is also copied to the plot file, using the terminal device definition. Note that this mode of operation is not standard and is NOT recommended for general use since it creates plot files that may be difficult to process with GPL, GPR or E2PLOTS. You can also use the Print Page option in EMTOOLW to print all the information actually on the screen. 5.2.8: How can I obtain a plot where node numbers (labels) are displayed for centroids only? Starting with Release 9, you can use the new sub-option 3 (several user defined link attributes) of option 6 (user defined data) to customize the display of link attributes, together with the specification of the subset of nodes for which the node number or label is displayed, by choosing option 3 of the following new Select question: Select: Type of node number display 1= display no node numbers 2= display all node numbers 3= display selected node numbers For more details, refer to the EMME/2 User's Manual, Module 2.13, option 6. For releases prior to Release 9, the cent2ann.awk script, available in the EMME/2, STAN Scripts section, can be used to generate an annotation file containing only the centroid numbers or labels, using the network input file (d211.in) as input. 5.3 Network and Matrix Calculators 5.3.1: How can I set color indices for different types of links? The best way is to use the network calculator (Module 2.41) to set the values in a link attribute. For example, if you want to color the connectors in green, the expression would simply be 3 (for color index 3) and, to apply the result to the connectors only, the link selection would be: ci=1 |cj=1 Note that the link attribute must first be initialized to the desired (positive) color for all other links since links with a color index of 0 are not displayed. 5.3.2: Can temporary link attributes tmplx be used in selections? No. However, it is possible to include conditions within the expression in order to "simulate" the selection. For example, if you want to change the length of the links only for those links having tmpl1 greater than 0, the expression would be: len * (tmpl1 <= 0) + expression * (tmpl1 > 0) It is also possible to replace temporary attributes by extra attributes, which can be used in the selections. 5.3.3: How can I write an expression that ensures a negative value is replaced by 0? Use the .max. operator: (expression) .max.0. 5.3.4: In the network and matrix calculators, why are the results saved when I send the report to the printer but not when I send the report to the terminal? Probably because you stopped the display of the report prematurely with a r or q. If you do not want to look at the remainder of the report, type l to skip to the end of the report. 5.4 Matrices 5.4.1: Why do some of my matrices have a "/c" flag? Probably because a transit assignment did not end normally. During a transit assignment, matrices are transformed to columnwise format in order to work destination by destination. At the end of the assignment, the matrices are retransformed to rowwise format. If the transit assignment does not end normally, the matrices will stay columnwise. Since working with columnwise matrices will increase the matrix access time drastically, you should change the storage type of columnwise matrices by using Module 3.12, option 3 (modify a matrix). 5.4.2: Matrix calculations are very slow. Is there a way to speed this up? You could do the following: Check that the matrices used in the calculations are not columnwise (see also Question 5.4.1). When running a macro with many calculations, suppress programmed pauses (~o|32). On Intel platforms, try to perform disk defragmentation. If you are using DOS or Windows 3.x, verify the size of your disk cache (Smartdrive). On UNIX platforms, try to use the data bank locally and not via a network. 5.4.3: How can I list several origin and destination matrices, excluding the zones for which the value is 0 for all matrices? Prepare an origin matrix that will be used as a constraint matrix. This matrix should contain the sum of all origin and destination matrices (destination matrices should be transposed!). For example: mo99 = mo1 + mo2 + md3'. If matrices contain negative values, the expression should be: abs(mo1) + abs(mo2) + abs(md3'). 5.5 Functions 5.5.1: How can I try different values for a constant that appears in my volume-delay function without entering the complete function each time? Put the value in a scalar matrix n and then refer to it in the function using ms(n). 5.5.2: One of the sample data files provided with the EMME/2 system contains functions. Are they general enough to be used with any data bank? No. These functions are specific to the Winnipeg data and are included solely for demonstration purposes. 5.6 Assignments 5.6.1: After an auto assignment, are the times on the connectors included in the auto time saved as a matrix? Yes. The time will depend on the functions for the connectors. 5.6.2: Is it possible to stop an auto assignment before one of the specified stopping criteria is met? Yes. Use Ctrl-C. The assignment will stop at the end of the next iteration. 5.6.3: How can I identify the non-assigned trips after an assignment? You can output the demand matrix using the time matrix as a constraint matrix with the constraint interval: (0.001,9999,exclude). This will include the trips on the diagonal (travel time of zero) and the non-connected trips (travel time equal to infinity). 5.6.4: I have a travel time matrix. How can I obtain the minimum travel time for each origin? In Module 3.21, perform a matrix computation and save the result in an origin matrix. Use the travel time matrix for the expression and as a constraint matrix (to exclude 0 values). Use the .min. operator for aggregation over destinations. 5.6.5: How can I perform an additional options auto assignment AND consider transit vehicles as additional volumes at the same time? You can compute the auto equivalent of the transit vehicles in Module 2.41 (vauteq*60/hdwy) and store the results in link and turn user data. Then, replace volad by ulN in the volume-delay functions, and pvolad by upN in the penalty functions. 5.6.6: How can I optimize the speed of my assignment? You can suppress the computation of the objective function by setting the relative gap to a negative value (for example, -1). You should also simplify your function expressions as much as possible. For example, replace sub-expressions such as constantA/constantB by constantC and use put() and get() functions whenever possible. 5.7 Zone Systems 5.7.1: Can I use different zone systems in the same data bank? It is possible but not recommended. It is better to have different data banks for different zone systems. Refer to the EMME/2 User's Manual, Chapter 3, Section 3.2.2 Matrices and scenarios. 5.7.2: I need more detailed zones in some parts of my network and more aggregate data in other parts. What is the best way to modify my zone system? Refer to the EMME/2 News of March 87 - Moving matrix data between different zone systems. The basic idea is to: create 2 data banks, 1 for each zone system in each data bank, create an ensemble that will be used to define the correspondence between the 2 zone systems punch and read the matrices using this ensemble (further manipulations might be required) 5.7.3: How can I increase the number of centroids in my data bank? While it is possible to increase most of the data bank dimensions by using the macro modifydb, it is not possible to increase the number of centroids. If this is necessary, the only solution is to create a new data bank with the new dimensions. Data from the previous data bank can then be imported using Module 1.31.
Normally, this information is not sent to the plot file. However, if switch 14 is ON, all graphic output which is transmitted to the terminal is also copied to the plot file, using the terminal device definition. Note that this mode of operation is not standard and is NOT recommended for general use since it creates plot files that may be difficult to process with GPL, GPR or E2PLOTS. You can also use the Print Page option in EMTOOLW to print all the information actually on the screen.
5.2.8: How can I obtain a plot where node numbers (labels) are displayed for centroids only? Starting with Release 9, you can use the new sub-option 3 (several user defined link attributes) of option 6 (user defined data) to customize the display of link attributes, together with the specification of the subset of nodes for which the node number or label is displayed, by choosing option 3 of the following new Select question: Select: Type of node number display 1= display no node numbers 2= display all node numbers 3= display selected node numbers For more details, refer to the EMME/2 User's Manual, Module 2.13, option 6. For releases prior to Release 9, the cent2ann.awk script, available in the EMME/2, STAN Scripts section, can be used to generate an annotation file containing only the centroid numbers or labels, using the network input file (d211.in) as input. 5.3 Network and Matrix Calculators 5.3.1: How can I set color indices for different types of links? The best way is to use the network calculator (Module 2.41) to set the values in a link attribute. For example, if you want to color the connectors in green, the expression would simply be 3 (for color index 3) and, to apply the result to the connectors only, the link selection would be: ci=1 |cj=1 Note that the link attribute must first be initialized to the desired (positive) color for all other links since links with a color index of 0 are not displayed. 5.3.2: Can temporary link attributes tmplx be used in selections? No. However, it is possible to include conditions within the expression in order to "simulate" the selection. For example, if you want to change the length of the links only for those links having tmpl1 greater than 0, the expression would be: len * (tmpl1 <= 0) + expression * (tmpl1 > 0) It is also possible to replace temporary attributes by extra attributes, which can be used in the selections. 5.3.3: How can I write an expression that ensures a negative value is replaced by 0? Use the .max. operator: (expression) .max.0. 5.3.4: In the network and matrix calculators, why are the results saved when I send the report to the printer but not when I send the report to the terminal? Probably because you stopped the display of the report prematurely with a r or q. If you do not want to look at the remainder of the report, type l to skip to the end of the report. 5.4 Matrices 5.4.1: Why do some of my matrices have a "/c" flag? Probably because a transit assignment did not end normally. During a transit assignment, matrices are transformed to columnwise format in order to work destination by destination. At the end of the assignment, the matrices are retransformed to rowwise format. If the transit assignment does not end normally, the matrices will stay columnwise. Since working with columnwise matrices will increase the matrix access time drastically, you should change the storage type of columnwise matrices by using Module 3.12, option 3 (modify a matrix). 5.4.2: Matrix calculations are very slow. Is there a way to speed this up? You could do the following: Check that the matrices used in the calculations are not columnwise (see also Question 5.4.1). When running a macro with many calculations, suppress programmed pauses (~o|32). On Intel platforms, try to perform disk defragmentation. If you are using DOS or Windows 3.x, verify the size of your disk cache (Smartdrive). On UNIX platforms, try to use the data bank locally and not via a network. 5.4.3: How can I list several origin and destination matrices, excluding the zones for which the value is 0 for all matrices? Prepare an origin matrix that will be used as a constraint matrix. This matrix should contain the sum of all origin and destination matrices (destination matrices should be transposed!). For example: mo99 = mo1 + mo2 + md3'. If matrices contain negative values, the expression should be: abs(mo1) + abs(mo2) + abs(md3'). 5.5 Functions 5.5.1: How can I try different values for a constant that appears in my volume-delay function without entering the complete function each time? Put the value in a scalar matrix n and then refer to it in the function using ms(n). 5.5.2: One of the sample data files provided with the EMME/2 system contains functions. Are they general enough to be used with any data bank? No. These functions are specific to the Winnipeg data and are included solely for demonstration purposes. 5.6 Assignments 5.6.1: After an auto assignment, are the times on the connectors included in the auto time saved as a matrix? Yes. The time will depend on the functions for the connectors. 5.6.2: Is it possible to stop an auto assignment before one of the specified stopping criteria is met? Yes. Use Ctrl-C. The assignment will stop at the end of the next iteration. 5.6.3: How can I identify the non-assigned trips after an assignment? You can output the demand matrix using the time matrix as a constraint matrix with the constraint interval: (0.001,9999,exclude). This will include the trips on the diagonal (travel time of zero) and the non-connected trips (travel time equal to infinity). 5.6.4: I have a travel time matrix. How can I obtain the minimum travel time for each origin? In Module 3.21, perform a matrix computation and save the result in an origin matrix. Use the travel time matrix for the expression and as a constraint matrix (to exclude 0 values). Use the .min. operator for aggregation over destinations. 5.6.5: How can I perform an additional options auto assignment AND consider transit vehicles as additional volumes at the same time? You can compute the auto equivalent of the transit vehicles in Module 2.41 (vauteq*60/hdwy) and store the results in link and turn user data. Then, replace volad by ulN in the volume-delay functions, and pvolad by upN in the penalty functions. 5.6.6: How can I optimize the speed of my assignment? You can suppress the computation of the objective function by setting the relative gap to a negative value (for example, -1). You should also simplify your function expressions as much as possible. For example, replace sub-expressions such as constantA/constantB by constantC and use put() and get() functions whenever possible. 5.7 Zone Systems 5.7.1: Can I use different zone systems in the same data bank? It is possible but not recommended. It is better to have different data banks for different zone systems. Refer to the EMME/2 User's Manual, Chapter 3, Section 3.2.2 Matrices and scenarios. 5.7.2: I need more detailed zones in some parts of my network and more aggregate data in other parts. What is the best way to modify my zone system? Refer to the EMME/2 News of March 87 - Moving matrix data between different zone systems. The basic idea is to: create 2 data banks, 1 for each zone system in each data bank, create an ensemble that will be used to define the correspondence between the 2 zone systems punch and read the matrices using this ensemble (further manipulations might be required) 5.7.3: How can I increase the number of centroids in my data bank? While it is possible to increase most of the data bank dimensions by using the macro modifydb, it is not possible to increase the number of centroids. If this is necessary, the only solution is to create a new data bank with the new dimensions. Data from the previous data bank can then be imported using Module 1.31.
Starting with Release 9, you can use the new sub-option 3 (several user defined link attributes) of option 6 (user defined data) to customize the display of link attributes, together with the specification of the subset of nodes for which the node number or label is displayed, by choosing option 3 of the following new Select question:
For more details, refer to the EMME/2 User's Manual, Module 2.13, option 6.
For releases prior to Release 9, the cent2ann.awk script, available in the EMME/2, STAN Scripts section, can be used to generate an annotation file containing only the centroid numbers or labels, using the network input file (d211.in) as input.
5.3 Network and Matrix Calculators 5.3.1: How can I set color indices for different types of links? The best way is to use the network calculator (Module 2.41) to set the values in a link attribute. For example, if you want to color the connectors in green, the expression would simply be 3 (for color index 3) and, to apply the result to the connectors only, the link selection would be: ci=1 |cj=1 Note that the link attribute must first be initialized to the desired (positive) color for all other links since links with a color index of 0 are not displayed. 5.3.2: Can temporary link attributes tmplx be used in selections? No. However, it is possible to include conditions within the expression in order to "simulate" the selection. For example, if you want to change the length of the links only for those links having tmpl1 greater than 0, the expression would be: len * (tmpl1 <= 0) + expression * (tmpl1 > 0) It is also possible to replace temporary attributes by extra attributes, which can be used in the selections. 5.3.3: How can I write an expression that ensures a negative value is replaced by 0? Use the .max. operator: (expression) .max.0. 5.3.4: In the network and matrix calculators, why are the results saved when I send the report to the printer but not when I send the report to the terminal? Probably because you stopped the display of the report prematurely with a r or q. If you do not want to look at the remainder of the report, type l to skip to the end of the report. 5.4 Matrices 5.4.1: Why do some of my matrices have a "/c" flag? Probably because a transit assignment did not end normally. During a transit assignment, matrices are transformed to columnwise format in order to work destination by destination. At the end of the assignment, the matrices are retransformed to rowwise format. If the transit assignment does not end normally, the matrices will stay columnwise. Since working with columnwise matrices will increase the matrix access time drastically, you should change the storage type of columnwise matrices by using Module 3.12, option 3 (modify a matrix). 5.4.2: Matrix calculations are very slow. Is there a way to speed this up? You could do the following: Check that the matrices used in the calculations are not columnwise (see also Question 5.4.1). When running a macro with many calculations, suppress programmed pauses (~o|32). On Intel platforms, try to perform disk defragmentation. If you are using DOS or Windows 3.x, verify the size of your disk cache (Smartdrive). On UNIX platforms, try to use the data bank locally and not via a network. 5.4.3: How can I list several origin and destination matrices, excluding the zones for which the value is 0 for all matrices? Prepare an origin matrix that will be used as a constraint matrix. This matrix should contain the sum of all origin and destination matrices (destination matrices should be transposed!). For example: mo99 = mo1 + mo2 + md3'. If matrices contain negative values, the expression should be: abs(mo1) + abs(mo2) + abs(md3'). 5.5 Functions 5.5.1: How can I try different values for a constant that appears in my volume-delay function without entering the complete function each time? Put the value in a scalar matrix n and then refer to it in the function using ms(n). 5.5.2: One of the sample data files provided with the EMME/2 system contains functions. Are they general enough to be used with any data bank? No. These functions are specific to the Winnipeg data and are included solely for demonstration purposes. 5.6 Assignments 5.6.1: After an auto assignment, are the times on the connectors included in the auto time saved as a matrix? Yes. The time will depend on the functions for the connectors. 5.6.2: Is it possible to stop an auto assignment before one of the specified stopping criteria is met? Yes. Use Ctrl-C. The assignment will stop at the end of the next iteration. 5.6.3: How can I identify the non-assigned trips after an assignment? You can output the demand matrix using the time matrix as a constraint matrix with the constraint interval: (0.001,9999,exclude). This will include the trips on the diagonal (travel time of zero) and the non-connected trips (travel time equal to infinity). 5.6.4: I have a travel time matrix. How can I obtain the minimum travel time for each origin? In Module 3.21, perform a matrix computation and save the result in an origin matrix. Use the travel time matrix for the expression and as a constraint matrix (to exclude 0 values). Use the .min. operator for aggregation over destinations. 5.6.5: How can I perform an additional options auto assignment AND consider transit vehicles as additional volumes at the same time? You can compute the auto equivalent of the transit vehicles in Module 2.41 (vauteq*60/hdwy) and store the results in link and turn user data. Then, replace volad by ulN in the volume-delay functions, and pvolad by upN in the penalty functions. 5.6.6: How can I optimize the speed of my assignment? You can suppress the computation of the objective function by setting the relative gap to a negative value (for example, -1). You should also simplify your function expressions as much as possible. For example, replace sub-expressions such as constantA/constantB by constantC and use put() and get() functions whenever possible. 5.7 Zone Systems 5.7.1: Can I use different zone systems in the same data bank? It is possible but not recommended. It is better to have different data banks for different zone systems. Refer to the EMME/2 User's Manual, Chapter 3, Section 3.2.2 Matrices and scenarios. 5.7.2: I need more detailed zones in some parts of my network and more aggregate data in other parts. What is the best way to modify my zone system? Refer to the EMME/2 News of March 87 - Moving matrix data between different zone systems. The basic idea is to: create 2 data banks, 1 for each zone system in each data bank, create an ensemble that will be used to define the correspondence between the 2 zone systems punch and read the matrices using this ensemble (further manipulations might be required) 5.7.3: How can I increase the number of centroids in my data bank? While it is possible to increase most of the data bank dimensions by using the macro modifydb, it is not possible to increase the number of centroids. If this is necessary, the only solution is to create a new data bank with the new dimensions. Data from the previous data bank can then be imported using Module 1.31.
5.3.1: How can I set color indices for different types of links? The best way is to use the network calculator (Module 2.41) to set the values in a link attribute. For example, if you want to color the connectors in green, the expression would simply be 3 (for color index 3) and, to apply the result to the connectors only, the link selection would be: ci=1 |cj=1 Note that the link attribute must first be initialized to the desired (positive) color for all other links since links with a color index of 0 are not displayed. 5.3.2: Can temporary link attributes tmplx be used in selections? No. However, it is possible to include conditions within the expression in order to "simulate" the selection. For example, if you want to change the length of the links only for those links having tmpl1 greater than 0, the expression would be: len * (tmpl1 <= 0) + expression * (tmpl1 > 0) It is also possible to replace temporary attributes by extra attributes, which can be used in the selections. 5.3.3: How can I write an expression that ensures a negative value is replaced by 0? Use the .max. operator: (expression) .max.0. 5.3.4: In the network and matrix calculators, why are the results saved when I send the report to the printer but not when I send the report to the terminal? Probably because you stopped the display of the report prematurely with a r or q. If you do not want to look at the remainder of the report, type l to skip to the end of the report. 5.4 Matrices 5.4.1: Why do some of my matrices have a "/c" flag? Probably because a transit assignment did not end normally. During a transit assignment, matrices are transformed to columnwise format in order to work destination by destination. At the end of the assignment, the matrices are retransformed to rowwise format. If the transit assignment does not end normally, the matrices will stay columnwise. Since working with columnwise matrices will increase the matrix access time drastically, you should change the storage type of columnwise matrices by using Module 3.12, option 3 (modify a matrix). 5.4.2: Matrix calculations are very slow. Is there a way to speed this up? You could do the following: Check that the matrices used in the calculations are not columnwise (see also Question 5.4.1). When running a macro with many calculations, suppress programmed pauses (~o|32). On Intel platforms, try to perform disk defragmentation. If you are using DOS or Windows 3.x, verify the size of your disk cache (Smartdrive). On UNIX platforms, try to use the data bank locally and not via a network. 5.4.3: How can I list several origin and destination matrices, excluding the zones for which the value is 0 for all matrices? Prepare an origin matrix that will be used as a constraint matrix. This matrix should contain the sum of all origin and destination matrices (destination matrices should be transposed!). For example: mo99 = mo1 + mo2 + md3'. If matrices contain negative values, the expression should be: abs(mo1) + abs(mo2) + abs(md3'). 5.5 Functions 5.5.1: How can I try different values for a constant that appears in my volume-delay function without entering the complete function each time? Put the value in a scalar matrix n and then refer to it in the function using ms(n). 5.5.2: One of the sample data files provided with the EMME/2 system contains functions. Are they general enough to be used with any data bank? No. These functions are specific to the Winnipeg data and are included solely for demonstration purposes. 5.6 Assignments 5.6.1: After an auto assignment, are the times on the connectors included in the auto time saved as a matrix? Yes. The time will depend on the functions for the connectors. 5.6.2: Is it possible to stop an auto assignment before one of the specified stopping criteria is met? Yes. Use Ctrl-C. The assignment will stop at the end of the next iteration. 5.6.3: How can I identify the non-assigned trips after an assignment? You can output the demand matrix using the time matrix as a constraint matrix with the constraint interval: (0.001,9999,exclude). This will include the trips on the diagonal (travel time of zero) and the non-connected trips (travel time equal to infinity). 5.6.4: I have a travel time matrix. How can I obtain the minimum travel time for each origin? In Module 3.21, perform a matrix computation and save the result in an origin matrix. Use the travel time matrix for the expression and as a constraint matrix (to exclude 0 values). Use the .min. operator for aggregation over destinations. 5.6.5: How can I perform an additional options auto assignment AND consider transit vehicles as additional volumes at the same time? You can compute the auto equivalent of the transit vehicles in Module 2.41 (vauteq*60/hdwy) and store the results in link and turn user data. Then, replace volad by ulN in the volume-delay functions, and pvolad by upN in the penalty functions. 5.6.6: How can I optimize the speed of my assignment? You can suppress the computation of the objective function by setting the relative gap to a negative value (for example, -1). You should also simplify your function expressions as much as possible. For example, replace sub-expressions such as constantA/constantB by constantC and use put() and get() functions whenever possible. 5.7 Zone Systems 5.7.1: Can I use different zone systems in the same data bank? It is possible but not recommended. It is better to have different data banks for different zone systems. Refer to the EMME/2 User's Manual, Chapter 3, Section 3.2.2 Matrices and scenarios. 5.7.2: I need more detailed zones in some parts of my network and more aggregate data in other parts. What is the best way to modify my zone system? Refer to the EMME/2 News of March 87 - Moving matrix data between different zone systems. The basic idea is to: create 2 data banks, 1 for each zone system in each data bank, create an ensemble that will be used to define the correspondence between the 2 zone systems punch and read the matrices using this ensemble (further manipulations might be required) 5.7.3: How can I increase the number of centroids in my data bank? While it is possible to increase most of the data bank dimensions by using the macro modifydb, it is not possible to increase the number of centroids. If this is necessary, the only solution is to create a new data bank with the new dimensions. Data from the previous data bank can then be imported using Module 1.31.
The best way is to use the network calculator (Module 2.41) to set the values in a link attribute. For example, if you want to color the connectors in green, the expression would simply be 3 (for color index 3) and, to apply the result to the connectors only, the link selection would be:
Note that the link attribute must first be initialized to the desired (positive) color for all other links since links with a color index of 0 are not displayed.
5.3.2: Can temporary link attributes tmplx be used in selections? No. However, it is possible to include conditions within the expression in order to "simulate" the selection. For example, if you want to change the length of the links only for those links having tmpl1 greater than 0, the expression would be: len * (tmpl1 <= 0) + expression * (tmpl1 > 0) It is also possible to replace temporary attributes by extra attributes, which can be used in the selections. 5.3.3: How can I write an expression that ensures a negative value is replaced by 0? Use the .max. operator: (expression) .max.0. 5.3.4: In the network and matrix calculators, why are the results saved when I send the report to the printer but not when I send the report to the terminal? Probably because you stopped the display of the report prematurely with a r or q. If you do not want to look at the remainder of the report, type l to skip to the end of the report. 5.4 Matrices 5.4.1: Why do some of my matrices have a "/c" flag? Probably because a transit assignment did not end normally. During a transit assignment, matrices are transformed to columnwise format in order to work destination by destination. At the end of the assignment, the matrices are retransformed to rowwise format. If the transit assignment does not end normally, the matrices will stay columnwise. Since working with columnwise matrices will increase the matrix access time drastically, you should change the storage type of columnwise matrices by using Module 3.12, option 3 (modify a matrix). 5.4.2: Matrix calculations are very slow. Is there a way to speed this up? You could do the following: Check that the matrices used in the calculations are not columnwise (see also Question 5.4.1). When running a macro with many calculations, suppress programmed pauses (~o|32). On Intel platforms, try to perform disk defragmentation. If you are using DOS or Windows 3.x, verify the size of your disk cache (Smartdrive). On UNIX platforms, try to use the data bank locally and not via a network. 5.4.3: How can I list several origin and destination matrices, excluding the zones for which the value is 0 for all matrices? Prepare an origin matrix that will be used as a constraint matrix. This matrix should contain the sum of all origin and destination matrices (destination matrices should be transposed!). For example: mo99 = mo1 + mo2 + md3'. If matrices contain negative values, the expression should be: abs(mo1) + abs(mo2) + abs(md3'). 5.5 Functions 5.5.1: How can I try different values for a constant that appears in my volume-delay function without entering the complete function each time? Put the value in a scalar matrix n and then refer to it in the function using ms(n). 5.5.2: One of the sample data files provided with the EMME/2 system contains functions. Are they general enough to be used with any data bank? No. These functions are specific to the Winnipeg data and are included solely for demonstration purposes. 5.6 Assignments 5.6.1: After an auto assignment, are the times on the connectors included in the auto time saved as a matrix? Yes. The time will depend on the functions for the connectors. 5.6.2: Is it possible to stop an auto assignment before one of the specified stopping criteria is met? Yes. Use Ctrl-C. The assignment will stop at the end of the next iteration. 5.6.3: How can I identify the non-assigned trips after an assignment? You can output the demand matrix using the time matrix as a constraint matrix with the constraint interval: (0.001,9999,exclude). This will include the trips on the diagonal (travel time of zero) and the non-connected trips (travel time equal to infinity). 5.6.4: I have a travel time matrix. How can I obtain the minimum travel time for each origin? In Module 3.21, perform a matrix computation and save the result in an origin matrix. Use the travel time matrix for the expression and as a constraint matrix (to exclude 0 values). Use the .min. operator for aggregation over destinations. 5.6.5: How can I perform an additional options auto assignment AND consider transit vehicles as additional volumes at the same time? You can compute the auto equivalent of the transit vehicles in Module 2.41 (vauteq*60/hdwy) and store the results in link and turn user data. Then, replace volad by ulN in the volume-delay functions, and pvolad by upN in the penalty functions. 5.6.6: How can I optimize the speed of my assignment? You can suppress the computation of the objective function by setting the relative gap to a negative value (for example, -1). You should also simplify your function expressions as much as possible. For example, replace sub-expressions such as constantA/constantB by constantC and use put() and get() functions whenever possible. 5.7 Zone Systems 5.7.1: Can I use different zone systems in the same data bank? It is possible but not recommended. It is better to have different data banks for different zone systems. Refer to the EMME/2 User's Manual, Chapter 3, Section 3.2.2 Matrices and scenarios. 5.7.2: I need more detailed zones in some parts of my network and more aggregate data in other parts. What is the best way to modify my zone system? Refer to the EMME/2 News of March 87 - Moving matrix data between different zone systems. The basic idea is to: create 2 data banks, 1 for each zone system in each data bank, create an ensemble that will be used to define the correspondence between the 2 zone systems punch and read the matrices using this ensemble (further manipulations might be required) 5.7.3: How can I increase the number of centroids in my data bank? While it is possible to increase most of the data bank dimensions by using the macro modifydb, it is not possible to increase the number of centroids. If this is necessary, the only solution is to create a new data bank with the new dimensions. Data from the previous data bank can then be imported using Module 1.31.
No. However, it is possible to include conditions within the expression in order to "simulate" the selection. For example, if you want to change the length of the links only for those links having tmpl1 greater than 0, the expression would be:
It is also possible to replace temporary attributes by extra attributes, which can be used in the selections.
5.3.3: How can I write an expression that ensures a negative value is replaced by 0? Use the .max. operator: (expression) .max.0. 5.3.4: In the network and matrix calculators, why are the results saved when I send the report to the printer but not when I send the report to the terminal? Probably because you stopped the display of the report prematurely with a r or q. If you do not want to look at the remainder of the report, type l to skip to the end of the report. 5.4 Matrices 5.4.1: Why do some of my matrices have a "/c" flag? Probably because a transit assignment did not end normally. During a transit assignment, matrices are transformed to columnwise format in order to work destination by destination. At the end of the assignment, the matrices are retransformed to rowwise format. If the transit assignment does not end normally, the matrices will stay columnwise. Since working with columnwise matrices will increase the matrix access time drastically, you should change the storage type of columnwise matrices by using Module 3.12, option 3 (modify a matrix). 5.4.2: Matrix calculations are very slow. Is there a way to speed this up? You could do the following: Check that the matrices used in the calculations are not columnwise (see also Question 5.4.1). When running a macro with many calculations, suppress programmed pauses (~o|32). On Intel platforms, try to perform disk defragmentation. If you are using DOS or Windows 3.x, verify the size of your disk cache (Smartdrive). On UNIX platforms, try to use the data bank locally and not via a network. 5.4.3: How can I list several origin and destination matrices, excluding the zones for which the value is 0 for all matrices? Prepare an origin matrix that will be used as a constraint matrix. This matrix should contain the sum of all origin and destination matrices (destination matrices should be transposed!). For example: mo99 = mo1 + mo2 + md3'. If matrices contain negative values, the expression should be: abs(mo1) + abs(mo2) + abs(md3'). 5.5 Functions 5.5.1: How can I try different values for a constant that appears in my volume-delay function without entering the complete function each time? Put the value in a scalar matrix n and then refer to it in the function using ms(n). 5.5.2: One of the sample data files provided with the EMME/2 system contains functions. Are they general enough to be used with any data bank? No. These functions are specific to the Winnipeg data and are included solely for demonstration purposes. 5.6 Assignments 5.6.1: After an auto assignment, are the times on the connectors included in the auto time saved as a matrix? Yes. The time will depend on the functions for the connectors. 5.6.2: Is it possible to stop an auto assignment before one of the specified stopping criteria is met? Yes. Use Ctrl-C. The assignment will stop at the end of the next iteration. 5.6.3: How can I identify the non-assigned trips after an assignment? You can output the demand matrix using the time matrix as a constraint matrix with the constraint interval: (0.001,9999,exclude). This will include the trips on the diagonal (travel time of zero) and the non-connected trips (travel time equal to infinity). 5.6.4: I have a travel time matrix. How can I obtain the minimum travel time for each origin? In Module 3.21, perform a matrix computation and save the result in an origin matrix. Use the travel time matrix for the expression and as a constraint matrix (to exclude 0 values). Use the .min. operator for aggregation over destinations. 5.6.5: How can I perform an additional options auto assignment AND consider transit vehicles as additional volumes at the same time? You can compute the auto equivalent of the transit vehicles in Module 2.41 (vauteq*60/hdwy) and store the results in link and turn user data. Then, replace volad by ulN in the volume-delay functions, and pvolad by upN in the penalty functions. 5.6.6: How can I optimize the speed of my assignment? You can suppress the computation of the objective function by setting the relative gap to a negative value (for example, -1). You should also simplify your function expressions as much as possible. For example, replace sub-expressions such as constantA/constantB by constantC and use put() and get() functions whenever possible. 5.7 Zone Systems 5.7.1: Can I use different zone systems in the same data bank? It is possible but not recommended. It is better to have different data banks for different zone systems. Refer to the EMME/2 User's Manual, Chapter 3, Section 3.2.2 Matrices and scenarios. 5.7.2: I need more detailed zones in some parts of my network and more aggregate data in other parts. What is the best way to modify my zone system? Refer to the EMME/2 News of March 87 - Moving matrix data between different zone systems. The basic idea is to: create 2 data banks, 1 for each zone system in each data bank, create an ensemble that will be used to define the correspondence between the 2 zone systems punch and read the matrices using this ensemble (further manipulations might be required) 5.7.3: How can I increase the number of centroids in my data bank? While it is possible to increase most of the data bank dimensions by using the macro modifydb, it is not possible to increase the number of centroids. If this is necessary, the only solution is to create a new data bank with the new dimensions. Data from the previous data bank can then be imported using Module 1.31.
Use the .max. operator: (expression) .max.0.
5.3.4: In the network and matrix calculators, why are the results saved when I send the report to the printer but not when I send the report to the terminal? Probably because you stopped the display of the report prematurely with a r or q. If you do not want to look at the remainder of the report, type l to skip to the end of the report. 5.4 Matrices 5.4.1: Why do some of my matrices have a "/c" flag? Probably because a transit assignment did not end normally. During a transit assignment, matrices are transformed to columnwise format in order to work destination by destination. At the end of the assignment, the matrices are retransformed to rowwise format. If the transit assignment does not end normally, the matrices will stay columnwise. Since working with columnwise matrices will increase the matrix access time drastically, you should change the storage type of columnwise matrices by using Module 3.12, option 3 (modify a matrix). 5.4.2: Matrix calculations are very slow. Is there a way to speed this up? You could do the following: Check that the matrices used in the calculations are not columnwise (see also Question 5.4.1). When running a macro with many calculations, suppress programmed pauses (~o|32). On Intel platforms, try to perform disk defragmentation. If you are using DOS or Windows 3.x, verify the size of your disk cache (Smartdrive). On UNIX platforms, try to use the data bank locally and not via a network. 5.4.3: How can I list several origin and destination matrices, excluding the zones for which the value is 0 for all matrices? Prepare an origin matrix that will be used as a constraint matrix. This matrix should contain the sum of all origin and destination matrices (destination matrices should be transposed!). For example: mo99 = mo1 + mo2 + md3'. If matrices contain negative values, the expression should be: abs(mo1) + abs(mo2) + abs(md3'). 5.5 Functions 5.5.1: How can I try different values for a constant that appears in my volume-delay function without entering the complete function each time? Put the value in a scalar matrix n and then refer to it in the function using ms(n). 5.5.2: One of the sample data files provided with the EMME/2 system contains functions. Are they general enough to be used with any data bank? No. These functions are specific to the Winnipeg data and are included solely for demonstration purposes. 5.6 Assignments 5.6.1: After an auto assignment, are the times on the connectors included in the auto time saved as a matrix? Yes. The time will depend on the functions for the connectors. 5.6.2: Is it possible to stop an auto assignment before one of the specified stopping criteria is met? Yes. Use Ctrl-C. The assignment will stop at the end of the next iteration. 5.6.3: How can I identify the non-assigned trips after an assignment? You can output the demand matrix using the time matrix as a constraint matrix with the constraint interval: (0.001,9999,exclude). This will include the trips on the diagonal (travel time of zero) and the non-connected trips (travel time equal to infinity). 5.6.4: I have a travel time matrix. How can I obtain the minimum travel time for each origin? In Module 3.21, perform a matrix computation and save the result in an origin matrix. Use the travel time matrix for the expression and as a constraint matrix (to exclude 0 values). Use the .min. operator for aggregation over destinations. 5.6.5: How can I perform an additional options auto assignment AND consider transit vehicles as additional volumes at the same time? You can compute the auto equivalent of the transit vehicles in Module 2.41 (vauteq*60/hdwy) and store the results in link and turn user data. Then, replace volad by ulN in the volume-delay functions, and pvolad by upN in the penalty functions. 5.6.6: How can I optimize the speed of my assignment? You can suppress the computation of the objective function by setting the relative gap to a negative value (for example, -1). You should also simplify your function expressions as much as possible. For example, replace sub-expressions such as constantA/constantB by constantC and use put() and get() functions whenever possible. 5.7 Zone Systems 5.7.1: Can I use different zone systems in the same data bank? It is possible but not recommended. It is better to have different data banks for different zone systems. Refer to the EMME/2 User's Manual, Chapter 3, Section 3.2.2 Matrices and scenarios. 5.7.2: I need more detailed zones in some parts of my network and more aggregate data in other parts. What is the best way to modify my zone system? Refer to the EMME/2 News of March 87 - Moving matrix data between different zone systems. The basic idea is to: create 2 data banks, 1 for each zone system in each data bank, create an ensemble that will be used to define the correspondence between the 2 zone systems punch and read the matrices using this ensemble (further manipulations might be required) 5.7.3: How can I increase the number of centroids in my data bank? While it is possible to increase most of the data bank dimensions by using the macro modifydb, it is not possible to increase the number of centroids. If this is necessary, the only solution is to create a new data bank with the new dimensions. Data from the previous data bank can then be imported using Module 1.31.
Probably because you stopped the display of the report prematurely with a r or q. If you do not want to look at the remainder of the report, type l to skip to the end of the report.
5.4 Matrices 5.4.1: Why do some of my matrices have a "/c" flag? Probably because a transit assignment did not end normally. During a transit assignment, matrices are transformed to columnwise format in order to work destination by destination. At the end of the assignment, the matrices are retransformed to rowwise format. If the transit assignment does not end normally, the matrices will stay columnwise. Since working with columnwise matrices will increase the matrix access time drastically, you should change the storage type of columnwise matrices by using Module 3.12, option 3 (modify a matrix). 5.4.2: Matrix calculations are very slow. Is there a way to speed this up? You could do the following: Check that the matrices used in the calculations are not columnwise (see also Question 5.4.1). When running a macro with many calculations, suppress programmed pauses (~o|32). On Intel platforms, try to perform disk defragmentation. If you are using DOS or Windows 3.x, verify the size of your disk cache (Smartdrive). On UNIX platforms, try to use the data bank locally and not via a network. 5.4.3: How can I list several origin and destination matrices, excluding the zones for which the value is 0 for all matrices? Prepare an origin matrix that will be used as a constraint matrix. This matrix should contain the sum of all origin and destination matrices (destination matrices should be transposed!). For example: mo99 = mo1 + mo2 + md3'. If matrices contain negative values, the expression should be: abs(mo1) + abs(mo2) + abs(md3'). 5.5 Functions 5.5.1: How can I try different values for a constant that appears in my volume-delay function without entering the complete function each time? Put the value in a scalar matrix n and then refer to it in the function using ms(n). 5.5.2: One of the sample data files provided with the EMME/2 system contains functions. Are they general enough to be used with any data bank? No. These functions are specific to the Winnipeg data and are included solely for demonstration purposes. 5.6 Assignments 5.6.1: After an auto assignment, are the times on the connectors included in the auto time saved as a matrix? Yes. The time will depend on the functions for the connectors. 5.6.2: Is it possible to stop an auto assignment before one of the specified stopping criteria is met? Yes. Use Ctrl-C. The assignment will stop at the end of the next iteration. 5.6.3: How can I identify the non-assigned trips after an assignment? You can output the demand matrix using the time matrix as a constraint matrix with the constraint interval: (0.001,9999,exclude). This will include the trips on the diagonal (travel time of zero) and the non-connected trips (travel time equal to infinity). 5.6.4: I have a travel time matrix. How can I obtain the minimum travel time for each origin? In Module 3.21, perform a matrix computation and save the result in an origin matrix. Use the travel time matrix for the expression and as a constraint matrix (to exclude 0 values). Use the .min. operator for aggregation over destinations. 5.6.5: How can I perform an additional options auto assignment AND consider transit vehicles as additional volumes at the same time? You can compute the auto equivalent of the transit vehicles in Module 2.41 (vauteq*60/hdwy) and store the results in link and turn user data. Then, replace volad by ulN in the volume-delay functions, and pvolad by upN in the penalty functions. 5.6.6: How can I optimize the speed of my assignment? You can suppress the computation of the objective function by setting the relative gap to a negative value (for example, -1). You should also simplify your function expressions as much as possible. For example, replace sub-expressions such as constantA/constantB by constantC and use put() and get() functions whenever possible. 5.7 Zone Systems 5.7.1: Can I use different zone systems in the same data bank? It is possible but not recommended. It is better to have different data banks for different zone systems. Refer to the EMME/2 User's Manual, Chapter 3, Section 3.2.2 Matrices and scenarios. 5.7.2: I need more detailed zones in some parts of my network and more aggregate data in other parts. What is the best way to modify my zone system? Refer to the EMME/2 News of March 87 - Moving matrix data between different zone systems. The basic idea is to: create 2 data banks, 1 for each zone system in each data bank, create an ensemble that will be used to define the correspondence between the 2 zone systems punch and read the matrices using this ensemble (further manipulations might be required) 5.7.3: How can I increase the number of centroids in my data bank? While it is possible to increase most of the data bank dimensions by using the macro modifydb, it is not possible to increase the number of centroids. If this is necessary, the only solution is to create a new data bank with the new dimensions. Data from the previous data bank can then be imported using Module 1.31.
5.4.1: Why do some of my matrices have a "/c" flag? Probably because a transit assignment did not end normally. During a transit assignment, matrices are transformed to columnwise format in order to work destination by destination. At the end of the assignment, the matrices are retransformed to rowwise format. If the transit assignment does not end normally, the matrices will stay columnwise. Since working with columnwise matrices will increase the matrix access time drastically, you should change the storage type of columnwise matrices by using Module 3.12, option 3 (modify a matrix). 5.4.2: Matrix calculations are very slow. Is there a way to speed this up? You could do the following: Check that the matrices used in the calculations are not columnwise (see also Question 5.4.1). When running a macro with many calculations, suppress programmed pauses (~o|32). On Intel platforms, try to perform disk defragmentation. If you are using DOS or Windows 3.x, verify the size of your disk cache (Smartdrive). On UNIX platforms, try to use the data bank locally and not via a network. 5.4.3: How can I list several origin and destination matrices, excluding the zones for which the value is 0 for all matrices? Prepare an origin matrix that will be used as a constraint matrix. This matrix should contain the sum of all origin and destination matrices (destination matrices should be transposed!). For example: mo99 = mo1 + mo2 + md3'. If matrices contain negative values, the expression should be: abs(mo1) + abs(mo2) + abs(md3'). 5.5 Functions 5.5.1: How can I try different values for a constant that appears in my volume-delay function without entering the complete function each time? Put the value in a scalar matrix n and then refer to it in the function using ms(n). 5.5.2: One of the sample data files provided with the EMME/2 system contains functions. Are they general enough to be used with any data bank? No. These functions are specific to the Winnipeg data and are included solely for demonstration purposes. 5.6 Assignments 5.6.1: After an auto assignment, are the times on the connectors included in the auto time saved as a matrix? Yes. The time will depend on the functions for the connectors. 5.6.2: Is it possible to stop an auto assignment before one of the specified stopping criteria is met? Yes. Use Ctrl-C. The assignment will stop at the end of the next iteration. 5.6.3: How can I identify the non-assigned trips after an assignment? You can output the demand matrix using the time matrix as a constraint matrix with the constraint interval: (0.001,9999,exclude). This will include the trips on the diagonal (travel time of zero) and the non-connected trips (travel time equal to infinity). 5.6.4: I have a travel time matrix. How can I obtain the minimum travel time for each origin? In Module 3.21, perform a matrix computation and save the result in an origin matrix. Use the travel time matrix for the expression and as a constraint matrix (to exclude 0 values). Use the .min. operator for aggregation over destinations. 5.6.5: How can I perform an additional options auto assignment AND consider transit vehicles as additional volumes at the same time? You can compute the auto equivalent of the transit vehicles in Module 2.41 (vauteq*60/hdwy) and store the results in link and turn user data. Then, replace volad by ulN in the volume-delay functions, and pvolad by upN in the penalty functions. 5.6.6: How can I optimize the speed of my assignment? You can suppress the computation of the objective function by setting the relative gap to a negative value (for example, -1). You should also simplify your function expressions as much as possible. For example, replace sub-expressions such as constantA/constantB by constantC and use put() and get() functions whenever possible. 5.7 Zone Systems 5.7.1: Can I use different zone systems in the same data bank? It is possible but not recommended. It is better to have different data banks for different zone systems. Refer to the EMME/2 User's Manual, Chapter 3, Section 3.2.2 Matrices and scenarios. 5.7.2: I need more detailed zones in some parts of my network and more aggregate data in other parts. What is the best way to modify my zone system? Refer to the EMME/2 News of March 87 - Moving matrix data between different zone systems. The basic idea is to: create 2 data banks, 1 for each zone system in each data bank, create an ensemble that will be used to define the correspondence between the 2 zone systems punch and read the matrices using this ensemble (further manipulations might be required) 5.7.3: How can I increase the number of centroids in my data bank? While it is possible to increase most of the data bank dimensions by using the macro modifydb, it is not possible to increase the number of centroids. If this is necessary, the only solution is to create a new data bank with the new dimensions. Data from the previous data bank can then be imported using Module 1.31.
Probably because a transit assignment did not end normally. During a transit assignment, matrices are transformed to columnwise format in order to work destination by destination. At the end of the assignment, the matrices are retransformed to rowwise format. If the transit assignment does not end normally, the matrices will stay columnwise. Since working with columnwise matrices will increase the matrix access time drastically, you should change the storage type of columnwise matrices by using Module 3.12, option 3 (modify a matrix).
5.4.2: Matrix calculations are very slow. Is there a way to speed this up? You could do the following: Check that the matrices used in the calculations are not columnwise (see also Question 5.4.1). When running a macro with many calculations, suppress programmed pauses (~o|32). On Intel platforms, try to perform disk defragmentation. If you are using DOS or Windows 3.x, verify the size of your disk cache (Smartdrive). On UNIX platforms, try to use the data bank locally and not via a network. 5.4.3: How can I list several origin and destination matrices, excluding the zones for which the value is 0 for all matrices? Prepare an origin matrix that will be used as a constraint matrix. This matrix should contain the sum of all origin and destination matrices (destination matrices should be transposed!). For example: mo99 = mo1 + mo2 + md3'. If matrices contain negative values, the expression should be: abs(mo1) + abs(mo2) + abs(md3'). 5.5 Functions 5.5.1: How can I try different values for a constant that appears in my volume-delay function without entering the complete function each time? Put the value in a scalar matrix n and then refer to it in the function using ms(n). 5.5.2: One of the sample data files provided with the EMME/2 system contains functions. Are they general enough to be used with any data bank? No. These functions are specific to the Winnipeg data and are included solely for demonstration purposes. 5.6 Assignments 5.6.1: After an auto assignment, are the times on the connectors included in the auto time saved as a matrix? Yes. The time will depend on the functions for the connectors. 5.6.2: Is it possible to stop an auto assignment before one of the specified stopping criteria is met? Yes. Use Ctrl-C. The assignment will stop at the end of the next iteration. 5.6.3: How can I identify the non-assigned trips after an assignment? You can output the demand matrix using the time matrix as a constraint matrix with the constraint interval: (0.001,9999,exclude). This will include the trips on the diagonal (travel time of zero) and the non-connected trips (travel time equal to infinity). 5.6.4: I have a travel time matrix. How can I obtain the minimum travel time for each origin? In Module 3.21, perform a matrix computation and save the result in an origin matrix. Use the travel time matrix for the expression and as a constraint matrix (to exclude 0 values). Use the .min. operator for aggregation over destinations. 5.6.5: How can I perform an additional options auto assignment AND consider transit vehicles as additional volumes at the same time? You can compute the auto equivalent of the transit vehicles in Module 2.41 (vauteq*60/hdwy) and store the results in link and turn user data. Then, replace volad by ulN in the volume-delay functions, and pvolad by upN in the penalty functions. 5.6.6: How can I optimize the speed of my assignment? You can suppress the computation of the objective function by setting the relative gap to a negative value (for example, -1). You should also simplify your function expressions as much as possible. For example, replace sub-expressions such as constantA/constantB by constantC and use put() and get() functions whenever possible. 5.7 Zone Systems 5.7.1: Can I use different zone systems in the same data bank? It is possible but not recommended. It is better to have different data banks for different zone systems. Refer to the EMME/2 User's Manual, Chapter 3, Section 3.2.2 Matrices and scenarios. 5.7.2: I need more detailed zones in some parts of my network and more aggregate data in other parts. What is the best way to modify my zone system? Refer to the EMME/2 News of March 87 - Moving matrix data between different zone systems. The basic idea is to: create 2 data banks, 1 for each zone system in each data bank, create an ensemble that will be used to define the correspondence between the 2 zone systems punch and read the matrices using this ensemble (further manipulations might be required) 5.7.3: How can I increase the number of centroids in my data bank? While it is possible to increase most of the data bank dimensions by using the macro modifydb, it is not possible to increase the number of centroids. If this is necessary, the only solution is to create a new data bank with the new dimensions. Data from the previous data bank can then be imported using Module 1.31.
You could do the following:
Prepare an origin matrix that will be used as a constraint matrix. This matrix should contain the sum of all origin and destination matrices (destination matrices should be transposed!). For example: mo99 = mo1 + mo2 + md3'.
If matrices contain negative values, the expression should be: abs(mo1) + abs(mo2) + abs(md3').
5.5 Functions 5.5.1: How can I try different values for a constant that appears in my volume-delay function without entering the complete function each time? Put the value in a scalar matrix n and then refer to it in the function using ms(n). 5.5.2: One of the sample data files provided with the EMME/2 system contains functions. Are they general enough to be used with any data bank? No. These functions are specific to the Winnipeg data and are included solely for demonstration purposes. 5.6 Assignments 5.6.1: After an auto assignment, are the times on the connectors included in the auto time saved as a matrix? Yes. The time will depend on the functions for the connectors. 5.6.2: Is it possible to stop an auto assignment before one of the specified stopping criteria is met? Yes. Use Ctrl-C. The assignment will stop at the end of the next iteration. 5.6.3: How can I identify the non-assigned trips after an assignment? You can output the demand matrix using the time matrix as a constraint matrix with the constraint interval: (0.001,9999,exclude). This will include the trips on the diagonal (travel time of zero) and the non-connected trips (travel time equal to infinity). 5.6.4: I have a travel time matrix. How can I obtain the minimum travel time for each origin? In Module 3.21, perform a matrix computation and save the result in an origin matrix. Use the travel time matrix for the expression and as a constraint matrix (to exclude 0 values). Use the .min. operator for aggregation over destinations. 5.6.5: How can I perform an additional options auto assignment AND consider transit vehicles as additional volumes at the same time? You can compute the auto equivalent of the transit vehicles in Module 2.41 (vauteq*60/hdwy) and store the results in link and turn user data. Then, replace volad by ulN in the volume-delay functions, and pvolad by upN in the penalty functions. 5.6.6: How can I optimize the speed of my assignment? You can suppress the computation of the objective function by setting the relative gap to a negative value (for example, -1). You should also simplify your function expressions as much as possible. For example, replace sub-expressions such as constantA/constantB by constantC and use put() and get() functions whenever possible. 5.7 Zone Systems 5.7.1: Can I use different zone systems in the same data bank? It is possible but not recommended. It is better to have different data banks for different zone systems. Refer to the EMME/2 User's Manual, Chapter 3, Section 3.2.2 Matrices and scenarios. 5.7.2: I need more detailed zones in some parts of my network and more aggregate data in other parts. What is the best way to modify my zone system? Refer to the EMME/2 News of March 87 - Moving matrix data between different zone systems. The basic idea is to: create 2 data banks, 1 for each zone system in each data bank, create an ensemble that will be used to define the correspondence between the 2 zone systems punch and read the matrices using this ensemble (further manipulations might be required) 5.7.3: How can I increase the number of centroids in my data bank? While it is possible to increase most of the data bank dimensions by using the macro modifydb, it is not possible to increase the number of centroids. If this is necessary, the only solution is to create a new data bank with the new dimensions. Data from the previous data bank can then be imported using Module 1.31.
5.5.1: How can I try different values for a constant that appears in my volume-delay function without entering the complete function each time? Put the value in a scalar matrix n and then refer to it in the function using ms(n). 5.5.2: One of the sample data files provided with the EMME/2 system contains functions. Are they general enough to be used with any data bank? No. These functions are specific to the Winnipeg data and are included solely for demonstration purposes. 5.6 Assignments 5.6.1: After an auto assignment, are the times on the connectors included in the auto time saved as a matrix? Yes. The time will depend on the functions for the connectors. 5.6.2: Is it possible to stop an auto assignment before one of the specified stopping criteria is met? Yes. Use Ctrl-C. The assignment will stop at the end of the next iteration. 5.6.3: How can I identify the non-assigned trips after an assignment? You can output the demand matrix using the time matrix as a constraint matrix with the constraint interval: (0.001,9999,exclude). This will include the trips on the diagonal (travel time of zero) and the non-connected trips (travel time equal to infinity). 5.6.4: I have a travel time matrix. How can I obtain the minimum travel time for each origin? In Module 3.21, perform a matrix computation and save the result in an origin matrix. Use the travel time matrix for the expression and as a constraint matrix (to exclude 0 values). Use the .min. operator for aggregation over destinations. 5.6.5: How can I perform an additional options auto assignment AND consider transit vehicles as additional volumes at the same time? You can compute the auto equivalent of the transit vehicles in Module 2.41 (vauteq*60/hdwy) and store the results in link and turn user data. Then, replace volad by ulN in the volume-delay functions, and pvolad by upN in the penalty functions. 5.6.6: How can I optimize the speed of my assignment? You can suppress the computation of the objective function by setting the relative gap to a negative value (for example, -1). You should also simplify your function expressions as much as possible. For example, replace sub-expressions such as constantA/constantB by constantC and use put() and get() functions whenever possible. 5.7 Zone Systems 5.7.1: Can I use different zone systems in the same data bank? It is possible but not recommended. It is better to have different data banks for different zone systems. Refer to the EMME/2 User's Manual, Chapter 3, Section 3.2.2 Matrices and scenarios. 5.7.2: I need more detailed zones in some parts of my network and more aggregate data in other parts. What is the best way to modify my zone system? Refer to the EMME/2 News of March 87 - Moving matrix data between different zone systems. The basic idea is to: create 2 data banks, 1 for each zone system in each data bank, create an ensemble that will be used to define the correspondence between the 2 zone systems punch and read the matrices using this ensemble (further manipulations might be required) 5.7.3: How can I increase the number of centroids in my data bank? While it is possible to increase most of the data bank dimensions by using the macro modifydb, it is not possible to increase the number of centroids. If this is necessary, the only solution is to create a new data bank with the new dimensions. Data from the previous data bank can then be imported using Module 1.31.
Put the value in a scalar matrix n and then refer to it in the function using ms(n).
5.5.2: One of the sample data files provided with the EMME/2 system contains functions. Are they general enough to be used with any data bank? No. These functions are specific to the Winnipeg data and are included solely for demonstration purposes. 5.6 Assignments 5.6.1: After an auto assignment, are the times on the connectors included in the auto time saved as a matrix? Yes. The time will depend on the functions for the connectors. 5.6.2: Is it possible to stop an auto assignment before one of the specified stopping criteria is met? Yes. Use Ctrl-C. The assignment will stop at the end of the next iteration. 5.6.3: How can I identify the non-assigned trips after an assignment? You can output the demand matrix using the time matrix as a constraint matrix with the constraint interval: (0.001,9999,exclude). This will include the trips on the diagonal (travel time of zero) and the non-connected trips (travel time equal to infinity). 5.6.4: I have a travel time matrix. How can I obtain the minimum travel time for each origin? In Module 3.21, perform a matrix computation and save the result in an origin matrix. Use the travel time matrix for the expression and as a constraint matrix (to exclude 0 values). Use the .min. operator for aggregation over destinations. 5.6.5: How can I perform an additional options auto assignment AND consider transit vehicles as additional volumes at the same time? You can compute the auto equivalent of the transit vehicles in Module 2.41 (vauteq*60/hdwy) and store the results in link and turn user data. Then, replace volad by ulN in the volume-delay functions, and pvolad by upN in the penalty functions. 5.6.6: How can I optimize the speed of my assignment? You can suppress the computation of the objective function by setting the relative gap to a negative value (for example, -1). You should also simplify your function expressions as much as possible. For example, replace sub-expressions such as constantA/constantB by constantC and use put() and get() functions whenever possible. 5.7 Zone Systems 5.7.1: Can I use different zone systems in the same data bank? It is possible but not recommended. It is better to have different data banks for different zone systems. Refer to the EMME/2 User's Manual, Chapter 3, Section 3.2.2 Matrices and scenarios. 5.7.2: I need more detailed zones in some parts of my network and more aggregate data in other parts. What is the best way to modify my zone system? Refer to the EMME/2 News of March 87 - Moving matrix data between different zone systems. The basic idea is to: create 2 data banks, 1 for each zone system in each data bank, create an ensemble that will be used to define the correspondence between the 2 zone systems punch and read the matrices using this ensemble (further manipulations might be required) 5.7.3: How can I increase the number of centroids in my data bank? While it is possible to increase most of the data bank dimensions by using the macro modifydb, it is not possible to increase the number of centroids. If this is necessary, the only solution is to create a new data bank with the new dimensions. Data from the previous data bank can then be imported using Module 1.31.
No. These functions are specific to the Winnipeg data and are included solely for demonstration purposes.
5.6 Assignments 5.6.1: After an auto assignment, are the times on the connectors included in the auto time saved as a matrix? Yes. The time will depend on the functions for the connectors. 5.6.2: Is it possible to stop an auto assignment before one of the specified stopping criteria is met? Yes. Use Ctrl-C. The assignment will stop at the end of the next iteration. 5.6.3: How can I identify the non-assigned trips after an assignment? You can output the demand matrix using the time matrix as a constraint matrix with the constraint interval: (0.001,9999,exclude). This will include the trips on the diagonal (travel time of zero) and the non-connected trips (travel time equal to infinity). 5.6.4: I have a travel time matrix. How can I obtain the minimum travel time for each origin? In Module 3.21, perform a matrix computation and save the result in an origin matrix. Use the travel time matrix for the expression and as a constraint matrix (to exclude 0 values). Use the .min. operator for aggregation over destinations. 5.6.5: How can I perform an additional options auto assignment AND consider transit vehicles as additional volumes at the same time? You can compute the auto equivalent of the transit vehicles in Module 2.41 (vauteq*60/hdwy) and store the results in link and turn user data. Then, replace volad by ulN in the volume-delay functions, and pvolad by upN in the penalty functions. 5.6.6: How can I optimize the speed of my assignment? You can suppress the computation of the objective function by setting the relative gap to a negative value (for example, -1). You should also simplify your function expressions as much as possible. For example, replace sub-expressions such as constantA/constantB by constantC and use put() and get() functions whenever possible. 5.7 Zone Systems 5.7.1: Can I use different zone systems in the same data bank? It is possible but not recommended. It is better to have different data banks for different zone systems. Refer to the EMME/2 User's Manual, Chapter 3, Section 3.2.2 Matrices and scenarios. 5.7.2: I need more detailed zones in some parts of my network and more aggregate data in other parts. What is the best way to modify my zone system? Refer to the EMME/2 News of March 87 - Moving matrix data between different zone systems. The basic idea is to: create 2 data banks, 1 for each zone system in each data bank, create an ensemble that will be used to define the correspondence between the 2 zone systems punch and read the matrices using this ensemble (further manipulations might be required) 5.7.3: How can I increase the number of centroids in my data bank? While it is possible to increase most of the data bank dimensions by using the macro modifydb, it is not possible to increase the number of centroids. If this is necessary, the only solution is to create a new data bank with the new dimensions. Data from the previous data bank can then be imported using Module 1.31.
5.6.1: After an auto assignment, are the times on the connectors included in the auto time saved as a matrix? Yes. The time will depend on the functions for the connectors. 5.6.2: Is it possible to stop an auto assignment before one of the specified stopping criteria is met? Yes. Use Ctrl-C. The assignment will stop at the end of the next iteration. 5.6.3: How can I identify the non-assigned trips after an assignment? You can output the demand matrix using the time matrix as a constraint matrix with the constraint interval: (0.001,9999,exclude). This will include the trips on the diagonal (travel time of zero) and the non-connected trips (travel time equal to infinity). 5.6.4: I have a travel time matrix. How can I obtain the minimum travel time for each origin? In Module 3.21, perform a matrix computation and save the result in an origin matrix. Use the travel time matrix for the expression and as a constraint matrix (to exclude 0 values). Use the .min. operator for aggregation over destinations. 5.6.5: How can I perform an additional options auto assignment AND consider transit vehicles as additional volumes at the same time? You can compute the auto equivalent of the transit vehicles in Module 2.41 (vauteq*60/hdwy) and store the results in link and turn user data. Then, replace volad by ulN in the volume-delay functions, and pvolad by upN in the penalty functions. 5.6.6: How can I optimize the speed of my assignment? You can suppress the computation of the objective function by setting the relative gap to a negative value (for example, -1). You should also simplify your function expressions as much as possible. For example, replace sub-expressions such as constantA/constantB by constantC and use put() and get() functions whenever possible. 5.7 Zone Systems 5.7.1: Can I use different zone systems in the same data bank? It is possible but not recommended. It is better to have different data banks for different zone systems. Refer to the EMME/2 User's Manual, Chapter 3, Section 3.2.2 Matrices and scenarios. 5.7.2: I need more detailed zones in some parts of my network and more aggregate data in other parts. What is the best way to modify my zone system? Refer to the EMME/2 News of March 87 - Moving matrix data between different zone systems. The basic idea is to: create 2 data banks, 1 for each zone system in each data bank, create an ensemble that will be used to define the correspondence between the 2 zone systems punch and read the matrices using this ensemble (further manipulations might be required) 5.7.3: How can I increase the number of centroids in my data bank? While it is possible to increase most of the data bank dimensions by using the macro modifydb, it is not possible to increase the number of centroids. If this is necessary, the only solution is to create a new data bank with the new dimensions. Data from the previous data bank can then be imported using Module 1.31.
Yes. The time will depend on the functions for the connectors.
5.6.2: Is it possible to stop an auto assignment before one of the specified stopping criteria is met? Yes. Use Ctrl-C. The assignment will stop at the end of the next iteration. 5.6.3: How can I identify the non-assigned trips after an assignment? You can output the demand matrix using the time matrix as a constraint matrix with the constraint interval: (0.001,9999,exclude). This will include the trips on the diagonal (travel time of zero) and the non-connected trips (travel time equal to infinity). 5.6.4: I have a travel time matrix. How can I obtain the minimum travel time for each origin? In Module 3.21, perform a matrix computation and save the result in an origin matrix. Use the travel time matrix for the expression and as a constraint matrix (to exclude 0 values). Use the .min. operator for aggregation over destinations. 5.6.5: How can I perform an additional options auto assignment AND consider transit vehicles as additional volumes at the same time? You can compute the auto equivalent of the transit vehicles in Module 2.41 (vauteq*60/hdwy) and store the results in link and turn user data. Then, replace volad by ulN in the volume-delay functions, and pvolad by upN in the penalty functions. 5.6.6: How can I optimize the speed of my assignment? You can suppress the computation of the objective function by setting the relative gap to a negative value (for example, -1). You should also simplify your function expressions as much as possible. For example, replace sub-expressions such as constantA/constantB by constantC and use put() and get() functions whenever possible. 5.7 Zone Systems 5.7.1: Can I use different zone systems in the same data bank? It is possible but not recommended. It is better to have different data banks for different zone systems. Refer to the EMME/2 User's Manual, Chapter 3, Section 3.2.2 Matrices and scenarios. 5.7.2: I need more detailed zones in some parts of my network and more aggregate data in other parts. What is the best way to modify my zone system? Refer to the EMME/2 News of March 87 - Moving matrix data between different zone systems. The basic idea is to: create 2 data banks, 1 for each zone system in each data bank, create an ensemble that will be used to define the correspondence between the 2 zone systems punch and read the matrices using this ensemble (further manipulations might be required) 5.7.3: How can I increase the number of centroids in my data bank? While it is possible to increase most of the data bank dimensions by using the macro modifydb, it is not possible to increase the number of centroids. If this is necessary, the only solution is to create a new data bank with the new dimensions. Data from the previous data bank can then be imported using Module 1.31.
Yes. Use Ctrl-C. The assignment will stop at the end of the next iteration.
5.6.3: How can I identify the non-assigned trips after an assignment? You can output the demand matrix using the time matrix as a constraint matrix with the constraint interval: (0.001,9999,exclude). This will include the trips on the diagonal (travel time of zero) and the non-connected trips (travel time equal to infinity). 5.6.4: I have a travel time matrix. How can I obtain the minimum travel time for each origin? In Module 3.21, perform a matrix computation and save the result in an origin matrix. Use the travel time matrix for the expression and as a constraint matrix (to exclude 0 values). Use the .min. operator for aggregation over destinations. 5.6.5: How can I perform an additional options auto assignment AND consider transit vehicles as additional volumes at the same time? You can compute the auto equivalent of the transit vehicles in Module 2.41 (vauteq*60/hdwy) and store the results in link and turn user data. Then, replace volad by ulN in the volume-delay functions, and pvolad by upN in the penalty functions. 5.6.6: How can I optimize the speed of my assignment? You can suppress the computation of the objective function by setting the relative gap to a negative value (for example, -1). You should also simplify your function expressions as much as possible. For example, replace sub-expressions such as constantA/constantB by constantC and use put() and get() functions whenever possible. 5.7 Zone Systems 5.7.1: Can I use different zone systems in the same data bank? It is possible but not recommended. It is better to have different data banks for different zone systems. Refer to the EMME/2 User's Manual, Chapter 3, Section 3.2.2 Matrices and scenarios. 5.7.2: I need more detailed zones in some parts of my network and more aggregate data in other parts. What is the best way to modify my zone system? Refer to the EMME/2 News of March 87 - Moving matrix data between different zone systems. The basic idea is to: create 2 data banks, 1 for each zone system in each data bank, create an ensemble that will be used to define the correspondence between the 2 zone systems punch and read the matrices using this ensemble (further manipulations might be required) 5.7.3: How can I increase the number of centroids in my data bank? While it is possible to increase most of the data bank dimensions by using the macro modifydb, it is not possible to increase the number of centroids. If this is necessary, the only solution is to create a new data bank with the new dimensions. Data from the previous data bank can then be imported using Module 1.31.
You can output the demand matrix using the time matrix as a constraint matrix with the constraint interval: (0.001,9999,exclude). This will include the trips on the diagonal (travel time of zero) and the non-connected trips (travel time equal to infinity).
5.6.4: I have a travel time matrix. How can I obtain the minimum travel time for each origin? In Module 3.21, perform a matrix computation and save the result in an origin matrix. Use the travel time matrix for the expression and as a constraint matrix (to exclude 0 values). Use the .min. operator for aggregation over destinations. 5.6.5: How can I perform an additional options auto assignment AND consider transit vehicles as additional volumes at the same time? You can compute the auto equivalent of the transit vehicles in Module 2.41 (vauteq*60/hdwy) and store the results in link and turn user data. Then, replace volad by ulN in the volume-delay functions, and pvolad by upN in the penalty functions. 5.6.6: How can I optimize the speed of my assignment? You can suppress the computation of the objective function by setting the relative gap to a negative value (for example, -1). You should also simplify your function expressions as much as possible. For example, replace sub-expressions such as constantA/constantB by constantC and use put() and get() functions whenever possible. 5.7 Zone Systems 5.7.1: Can I use different zone systems in the same data bank? It is possible but not recommended. It is better to have different data banks for different zone systems. Refer to the EMME/2 User's Manual, Chapter 3, Section 3.2.2 Matrices and scenarios. 5.7.2: I need more detailed zones in some parts of my network and more aggregate data in other parts. What is the best way to modify my zone system? Refer to the EMME/2 News of March 87 - Moving matrix data between different zone systems. The basic idea is to: create 2 data banks, 1 for each zone system in each data bank, create an ensemble that will be used to define the correspondence between the 2 zone systems punch and read the matrices using this ensemble (further manipulations might be required) 5.7.3: How can I increase the number of centroids in my data bank? While it is possible to increase most of the data bank dimensions by using the macro modifydb, it is not possible to increase the number of centroids. If this is necessary, the only solution is to create a new data bank with the new dimensions. Data from the previous data bank can then be imported using Module 1.31.
In Module 3.21, perform a matrix computation and save the result in an origin matrix. Use the travel time matrix for the expression and as a constraint matrix (to exclude 0 values). Use the .min. operator for aggregation over destinations.
5.6.5: How can I perform an additional options auto assignment AND consider transit vehicles as additional volumes at the same time? You can compute the auto equivalent of the transit vehicles in Module 2.41 (vauteq*60/hdwy) and store the results in link and turn user data. Then, replace volad by ulN in the volume-delay functions, and pvolad by upN in the penalty functions. 5.6.6: How can I optimize the speed of my assignment? You can suppress the computation of the objective function by setting the relative gap to a negative value (for example, -1). You should also simplify your function expressions as much as possible. For example, replace sub-expressions such as constantA/constantB by constantC and use put() and get() functions whenever possible. 5.7 Zone Systems 5.7.1: Can I use different zone systems in the same data bank? It is possible but not recommended. It is better to have different data banks for different zone systems. Refer to the EMME/2 User's Manual, Chapter 3, Section 3.2.2 Matrices and scenarios. 5.7.2: I need more detailed zones in some parts of my network and more aggregate data in other parts. What is the best way to modify my zone system? Refer to the EMME/2 News of March 87 - Moving matrix data between different zone systems. The basic idea is to: create 2 data banks, 1 for each zone system in each data bank, create an ensemble that will be used to define the correspondence between the 2 zone systems punch and read the matrices using this ensemble (further manipulations might be required) 5.7.3: How can I increase the number of centroids in my data bank? While it is possible to increase most of the data bank dimensions by using the macro modifydb, it is not possible to increase the number of centroids. If this is necessary, the only solution is to create a new data bank with the new dimensions. Data from the previous data bank can then be imported using Module 1.31.
You can compute the auto equivalent of the transit vehicles in Module 2.41 (vauteq*60/hdwy) and store the results in link and turn user data. Then, replace volad by ulN in the volume-delay functions, and pvolad by upN in the penalty functions.
5.6.6: How can I optimize the speed of my assignment? You can suppress the computation of the objective function by setting the relative gap to a negative value (for example, -1). You should also simplify your function expressions as much as possible. For example, replace sub-expressions such as constantA/constantB by constantC and use put() and get() functions whenever possible. 5.7 Zone Systems 5.7.1: Can I use different zone systems in the same data bank? It is possible but not recommended. It is better to have different data banks for different zone systems. Refer to the EMME/2 User's Manual, Chapter 3, Section 3.2.2 Matrices and scenarios. 5.7.2: I need more detailed zones in some parts of my network and more aggregate data in other parts. What is the best way to modify my zone system? Refer to the EMME/2 News of March 87 - Moving matrix data between different zone systems. The basic idea is to: create 2 data banks, 1 for each zone system in each data bank, create an ensemble that will be used to define the correspondence between the 2 zone systems punch and read the matrices using this ensemble (further manipulations might be required) 5.7.3: How can I increase the number of centroids in my data bank? While it is possible to increase most of the data bank dimensions by using the macro modifydb, it is not possible to increase the number of centroids. If this is necessary, the only solution is to create a new data bank with the new dimensions. Data from the previous data bank can then be imported using Module 1.31.
You can suppress the computation of the objective function by setting the relative gap to a negative value (for example, -1). You should also simplify your function expressions as much as possible. For example, replace sub-expressions such as constantA/constantB by constantC and use put() and get() functions whenever possible. 5.7 Zone Systems 5.7.1: Can I use different zone systems in the same data bank? It is possible but not recommended. It is better to have different data banks for different zone systems. Refer to the EMME/2 User's Manual, Chapter 3, Section 3.2.2 Matrices and scenarios. 5.7.2: I need more detailed zones in some parts of my network and more aggregate data in other parts. What is the best way to modify my zone system? Refer to the EMME/2 News of March 87 - Moving matrix data between different zone systems. The basic idea is to: create 2 data banks, 1 for each zone system in each data bank, create an ensemble that will be used to define the correspondence between the 2 zone systems punch and read the matrices using this ensemble (further manipulations might be required) 5.7.3: How can I increase the number of centroids in my data bank? While it is possible to increase most of the data bank dimensions by using the macro modifydb, it is not possible to increase the number of centroids. If this is necessary, the only solution is to create a new data bank with the new dimensions. Data from the previous data bank can then be imported using Module 1.31.
5.7.1: Can I use different zone systems in the same data bank? It is possible but not recommended. It is better to have different data banks for different zone systems. Refer to the EMME/2 User's Manual, Chapter 3, Section 3.2.2 Matrices and scenarios. 5.7.2: I need more detailed zones in some parts of my network and more aggregate data in other parts. What is the best way to modify my zone system? Refer to the EMME/2 News of March 87 - Moving matrix data between different zone systems. The basic idea is to: create 2 data banks, 1 for each zone system in each data bank, create an ensemble that will be used to define the correspondence between the 2 zone systems punch and read the matrices using this ensemble (further manipulations might be required) 5.7.3: How can I increase the number of centroids in my data bank? While it is possible to increase most of the data bank dimensions by using the macro modifydb, it is not possible to increase the number of centroids. If this is necessary, the only solution is to create a new data bank with the new dimensions. Data from the previous data bank can then be imported using Module 1.31.
It is possible but not recommended. It is better to have different data banks for different zone systems. Refer to the EMME/2 User's Manual, Chapter 3, Section 3.2.2 Matrices and scenarios.
5.7.2: I need more detailed zones in some parts of my network and more aggregate data in other parts. What is the best way to modify my zone system? Refer to the EMME/2 News of March 87 - Moving matrix data between different zone systems. The basic idea is to: create 2 data banks, 1 for each zone system in each data bank, create an ensemble that will be used to define the correspondence between the 2 zone systems punch and read the matrices using this ensemble (further manipulations might be required) 5.7.3: How can I increase the number of centroids in my data bank? While it is possible to increase most of the data bank dimensions by using the macro modifydb, it is not possible to increase the number of centroids. If this is necessary, the only solution is to create a new data bank with the new dimensions. Data from the previous data bank can then be imported using Module 1.31.
Refer to the EMME/2 News of March 87 - Moving matrix data between different zone systems. The basic idea is to:
While it is possible to increase most of the data bank dimensions by using the macro modifydb, it is not possible to increase the number of centroids. If this is necessary, the only solution is to create a new data bank with the new dimensions. Data from the previous data bank can then be imported using Module 1.31.
上一篇：