'Least squares’ parameters
Least squares: Count values
Element |
Description |
Use only network objects with volume > 0 and counted value > 0 |
If you select this option, only network objects with a volume and count value above 0 will be used. |
Use default weights |
If the option is selected, the value 1/sqrt(count value) is used as weight for all count values. This weighting automatically leads to a balance between large and small count values. Note The setting also affects distribution. |
Counted volumes refer to active line route items only |
If you select this option, the PuT count values will refer to active line route items only. The volumes or boarding/alighting passengers will be included on the basis of the following rules:
Note This option is only available if you have selected a demand segment of the PuT mode. |
Zones |
Base on totals of &matrix rows and columns If you select this option, the values of the numerical (possibly user-defined) zone attributes will be used as count data. Note When determining the origin/destination demand, PuT boarding/alighting volumes may be used only if these values do not include transfers. Only active zones If count data is not provided for all of the zones, you may use this option to regard only the data of active zones (Using filters to set network objects active or passive and Setting network objects active/passive via the spatial selection). Row total Via this button you can select the zone attribute that contains the origin demand counts. Column total Via this button you can select the zone attribute that contains the destination demand counts. Weight Use this button to select an attribute in a separate window for the weighting. Attribute 1.0 is the default. |
Links |
Base on counted link volumes If you select this option, the values of the numerical (possibly user-defined) zone attributes will be used as count data. Only active links If count data is not provided for all links, you may use this option to consider the data of active links only (Using filters to set network objects active or passive and Setting network objects active/passive via the spatial selection) Volume Use this button to select the link attribute that codes the volume counts of the link in a separate window. With the dynamic variant, you can select the virtual subattribute <Analysis time intervals> for attributes with the analysis time interval subattribute. Count values are then used per analysis time interval. Weight Use this button to select an attribute for the weighting in a separate window. Attribute 1.0 is the default. |
Turns and main turns |
Base on counted turn and main turn volumes If you select this option, the values of the numerical (possibly user-defined) turn or main turn attributes will be used as count data. Note To use a user-defined attribute for main turns, you also have to create this attribute for regular turns. Only active turns and main turns If count data is not provided for all of the turns or main turns respectively, you may use this option to regard only the data of active turns or main turns (if a main node has been defined) (Using filters to set network objects active or passive and Setting network objects active/passive via the spatial selection). Volume Via this button, you may - in a separate window - select the turn attribute that codes the volume counts values of turns or main turns. With the dynamic variant, you can select the virtual subattribute <Analysis time intervals> for attributes with the analysis time interval subattribute. Count values are then used per analysis time interval. Weight Use this button to select an attribute in a separate window for the weighting. Attribute 1.0 is the default. |
Screenlines |
Base on volumes counted in the direction of the screenline If this option has been selected, the values of numerical (possibly user-defined) screenline attributes are used as count data. Thereby, only volume data for the direction of the screenline is considered. Only active screenlines If count data is not provided for all of the screenlines you can use this option to consider only the data of active screenlines (Using filters to set network objects active or passive and Setting network objects active/passive via the spatial selection). Volume Via this button you can select the screenline attribute that contains the volume counts of the screenline. With the dynamic variant, you can select the virtual subattribute <Analysis time intervals> for attributes with the analysis time interval subattribute. Count values are then used per analysis time interval. Weight Use this button to select an attribute in a separate window for the weighting. Attribute 1.0 is the default. |
Total traffic |
Base on counted total traffic (all demand segments) If this option has been selected, the counted total traffic will be used. Note The total traffic counted for all demand segments is used. Total traffic In the entry field, enter the initial counted total traffic value. Weight In the input field, enter a value for weighting. |
Least squares: Count values PrT
Note: The Count values PrT tab does not appear unless you select a demand segment of the PrT mode. |
Element |
Description |
Lane groups |
Base on counted lane group volumes Select this option to use numerical values of lanes as count data. Here several lanes are divided into lane groups if there are mixed lanes that allow for turns in several directions (Fundamentals: Signalized nodes). Note If a lane belonging to a group contains the value zero, a warning message is displayed. Only lane groups of active nodes If count data is not available for all nodes, select this option to consider the lane values of active nodes only (Using filters to set network objects active or passive and Setting network objects active/passive via the spatial selection). Volume Click this button to select the lane attribute that contains the volume counts of the lane group in a separate window. Weight Use this button to select an attribute in a separate window for the weighting. Attribute 1.0 is the default. |
Least squares: Count values PuT
Notes: The Count values PuT tab is not shown unless you select a demand segment of the PuT mode. Count data can only be used if there are PuT paths for the selected demand segments. |
Element |
Description |
PuT lines |
Base on counted line boardings If you select this option, the values of the numerical (possibly user-defined) line attributes will be used as count data. Only active lines If count data is not provided for all PuT lines, you may use this option to include only the data of active lines (Using filters to set network objects active or passive and Setting network objects active/passive via the spatial selection). Boardings Use this button to select the line attribute that contains the count data of boarding passengers. Weight Use this button to select an attribute in a separate window for the weighting. Attribute 1.0 is the default. Base on counted passenger kilometers If you select this option, the values of the numerical (possibly user-defined) line attributes will be used as count data. Only active lines If count data is not provided for all PuT lines, you may use this option to include only the data of active lines (Using filters to set network objects active or passive and Setting network objects active/passive via the spatial selection). Passenger kilometers Via this button, you call a separate window: select the line attribute that codes the counted passenger kilometers of the line. Weight Use this button to select an attribute in a separate window for the weighting. Attribute 1.0 is the default. |
Stop areas |
Base on counted boarding/alighting passengers at stop areas If you select this option, the values of numerical (possibly user-defined) stop area attributes will be used as count data. Only active stop areas If count data is not provided for all stop areas, you may use this option to include only the data of active stop areas (Using filters to set network objects active or passive and Setting network objects active/passive via the spatial selection). Boardings Via this button you can select the stop area attribute that contains the count data of boarding passengers. Alightings Via this button you can select the stop area attribute that contains the count data of alighting passengers. Weight Use this button to select an attribute of the network object in a separate window. Attribute 1.0 is the default. Transfers count as boarding/alighting passengers Use the drop-down list to specify whether transfers are counted as boarding and alighting passengers.
|
Least squares: Distribution <code of demand segment>
Note: A tab is displayed for each selected demand segment. You can adjust several settings for each demand segment. |
Element |
Description |
Base on skim data distribution |
If you select this option, the demand segment output will be adjusted to a given distribution. |
Consider only active OD pairs |
If this option is selected, only the active OD pairs will be taken into account when the result matrix is adjusted to a given distribution. This applies to both class limit determination and OD pair determination in the output matrix. Note The values in the Number of trips column depend on this setting. |
Skim matrix for classification |
Via this button, you call the Select skim matrix window to select a skim matrix for the classification of OD pairs (Selecting matrices). |
Classes and shares |
Preset class limits and derive shares from current matrix If this option is selected, the class limits will be defined according to the following option settings: Shares according to the current demand matrix The shares of the specified classes are based on the current demand matrix. Shares according to the matrix Via this button, you call a separate window for the selection of the demand matrix, from which the share by specified class will be derived (Selecting matrices). Preset class limits and shares By selecting this option you specify the class limits in the list Class limits and shares below. Interval type Use this option to specify the class for each threshold. Note This option is provided only if the option From interval file has been selected. |
Class limits and shares |
List of specified classes. You can add the classes manually or import them from an interval file for further editing. If a skim matrix for classification or a matrix is selected, the corresponding shares and trips are displayed here. From / To Class limits for the desired distribution Note Classes must not overlap. Share Percentage of the class (100 % = 1) Note These fields cannot be edited if you import the shares from an existing demand matrix. Cumul. share Accumulated share Number of trips Number of trips resulting from the class limits and the entered share of the current demand matrix Create Use this button to add a new class. Delete Use this button to delete the selected class. Read from file Use this button to delete all existing classes and import new classes and shares from an interval file *.att into the list. Note This option is provided only if the option From interval file has been selected. |
From interval file |
Select this option to use an interval file with classes and shares. Click the button to choose a file. Notes This option allows you to use data that is not calculated until during the procedure sequence run. Interval files have the file extension * .att . |
Weight of the class shares |
Weighting of the class shares regarding the trips in the classes |
Least squares: Parameters
Element |
Description |
Adjust only OD pairs where M > 0 |
If this option has been selected, only OD pairs with a value greater than 0 in the filter matrix will be corrected (Creating a filter matrix). Any other OD pair is fixed and stored in the new output matrix without changes. The active OD pairs are adjusted in such a way that, as a result, the differences between old and new assignment volumes are already equalized. Via this button, you call a separate window for the selection of the filter matrix (Selecting matrices). |
Number of iterations |
After the user-defined number of iterations the iteration process is canceled though convergence has not been reached yet. |
Objective function |
Ratio of weights for demand deviation relative to count weights The parameter defines the ratio of the sum of the weights for the distances between the corrected matrix and the initial matrix () and between volume and count values (). The following formula can be used to describe the relation: Factor Only if the factor is greater than 0, the difference of the corrected matrix and the initial matrix will be considered in the target function. In case of a factor of 1, the sum of the weights of all relevant OD pairs equals the sum of the weights of all relevant count values. A factor greater than 1 means that the weights for the distance between the matrix values have a stronger effect, i.e. the corrected matrix values will deviate as little as possible from the values of the initial matrix. A smaller factor leads to minor deviations from the count values. Weight all time intervals and OD pairs equally If this option has been selected, all OD pairs are weighted equally. With the dynamic variant, all time intervals are also weighted equally. Weight OD pairs differently If this option has been selected, relative weights apply to each OD pair, depending on the weighting matrix selected for the demand segment. The specified matrix defines a relative weight for each OD pair with which the demand deviation enters the objective function. Note This option is not available in the static variant. Use individual weights per time interval and OD pair If this option is selected, relative weights apply to the demand deviation in the objective function. For each demand segment, select a weight matrix for the demand deviation per OD pair and time interval. For more information, refer to the help text in the program. Note This option is not available in the dynamic variant. |
|
Calculate deviations from current demand If this option is selected, the deviations from the current demand matrix are calculated. Calculate deviations from historic demand If this option is selected, you can specify an older demand matrix for the deviation of the matrix values and a standard time series for the dynamic variant for each demand segment, against which the deviations of the corrected demand matrix for the objective function are calculated. This is useful if the matrix correction procedure is executed in a procedure loop (with repeated assignment) and the demand deviation in each iteration is to be calculated for the same initial demand. |
OD pairs without count values |
If link volumes are used and none of the routes of an OD pair uses any of the counted links, data for matrix updating is missing for this OD pair. Leave unchanged If this option has been checked, the given OD data is stored in the new output matrix. Scale with average correction factor If this option has been selected, the given OD pairs are projected according to the mean changes to all of the other OD relations. Tips If for example only a link corridor has been counted in an extensive network and you would like to update just the corridor-using OD pairs, we recommend to set only the links of the corridor to the active state and to check this option. If only some individual OD pairs of lower priority do not use any of the counted links, we recommend to uncheck this option, since these links should be updated according to the general data changes. |
Flow matrix |
You can specify how the demand segment path information of the previous assignment shall be used for matrix correction by demand segment. |
DSeg |
Demand segments selected for matrix correction Note Each demand segment refers to a demand matrix (Connecting demand matrices and demand segments). |
Open/Save |
Recalculate If this option is checked, the flow matrix is recalculated for this particular demand segment. Load from file If this option is checked, the values of a calculated flow matrix are read from file for this particular demand segment. Tip If you want to perform several calculation runs for a demand segment, using the same demand matrix, but different count data (unchanged set of turns, links, main turns, zones) and/or procedure parameters, we recommend that during the first run, you calculate and save the flow matrix. For all subsequent runs, you can then read in the flow matrix again. Note Using an old flow matrix saves computing time but is a dangerous source of error, because the program is often unable to identify an unsuitable flow matrix. Make sure that the following applies.
|
File name |
Use the respective button to open a window. Depending on your Open/Save setting, you can enter or select a path and file name for the flow matrix you want to open (Load from file) or save (Recalculate). |
Least squares: Output (static)
Element |
Description |
Demand segment code |
Demand segments selected for matrix correction Note Each demand segment refers to a demand matrix (Connecting demand matrices and demand segments). |
Save result |
To demand matrix of DSeg Select this option to replace the demand segment data of the existing demand matrix in Visum by the calculated results. To matrix If this option has been selected, the demand segment's corrected demand matrix is saved. |
Result demand matrix |
If option To matrix has been selected, click the button and select the matrix in which the result is to be stored (Selecting matrices). |
Decimal places |
Number of decimal places for the storage of matrix values |
Least squares: Result time series (dynamic)
Element |
Description |
Demand segment code |
Demand segments selected for matrix correction Note The corrected demand is written per demand segment to the matrices referenced in the time series elements of the standard time series set for the respective demand segment. |
Save result |
in standard time series of the DSeg Select this option to replace the demand segment data of the existing standard time series in Visum by the calculated results. in standard time series If this option has been selected, the demand segment's corrected standard time series is saved. |
Time series number |
If the result is saved in the standard time series, you can select a previously defined result time series of the type matrix time series for each demand segment into which the result is written. Note If no standard matrix time series is allocated to the corresponding demand segment, the field is highlighted in red. |
Note: In the directory C:\Users\Public\Documents\PTV Vision\PTV Visum 2024/Examples, you can find an example of use on this topic. The Dynamic Matrix Estimation - Public Transport example describes the use of dynamic matrix estimation combined with the timetable-based public transport assignment. |