Cube Voyagercitilabs-website-resources.s3.amazonaws.com/resources/RS_CubeVoyager-6... · Cube...

Cube Voyager Release SummaryCube Voyager Release Summary

Citilabs®

Cube Voyager Release Summary

This document describes features introduced in different versions of Cube Voyager. Sections include:

• Version 6.4.3

• Version 6.4.2

• Version 6.4.1

• Version 6.4.0

• Version 6.1.1

• Version 6.1.0 Service Pack 1

• Version 6.1.0

• Version 6.0.2

• Version 6.0.1

• Version 6.0.0

• Version 5.1.4

• Version 5.1.3

• Version 5.1.2

• Version 5.1.0

• Version 5.0.2

Copyright © 2007–2017 Citilabs, Inc. All rights reserved.Citilabs is a registered trademark of Citilabs, Inc. All other brand names and product names are trademarks, registered trademarks or trade names of their respective holders.

Citilabs has carefully reviewed the accuracy of this document, but shall not be held responsible for any omissions or errors that may appear. Information in this document is subject to change without notice.

Citilabs, Inc.1605 E. Plaza Drive, Ste. 102Tallahassee FL 32308 USAwww.citilabs.com

Version 6.4.3 October 3, 2017

http://www.citilabs.com


• Version 5.0

• Version 4.2

• Version 4.1

• Version 4.0

• Version 3.2

• Version 3.1

In addition, this document discusses “Technical support and software updates” on page 61.

2 Cube Voyager Release Summary

Version 6.4.3

Version 6.4.3• Pilot

Fix for issue where PILOT prints incorrect values for variables

Fix a problem with string array when the index is not an integer

Better error handling for incorrect SMTP server in SENDMAIL

• Matrix Fix for issue with reading DBI ASCII file with line lengths

greater than 10,000 characters Updated error handling when defining Z in a RECI input Fix for matrix crashing issue with one line IF statements Updated matrix multi-dimensional arrays to support

memory allocations of > 2GB on 64-bit builds. Increase the number of sort fields allowed to 10 for RECI Fix for XCHOICE producing NAN values in output matrix

• Network Fix for CROSSTAB not working with 32bit installation

• Public Transport Fix for PT crowding crash with ICP and SL output files Additional convergence criteria for PT Crowding based on

new statistics (RMSE, RDIFF) Fix for issue with PT being unable to properly create the

route file .RTE during the route enumeration step with large numbers of zones

Fix for issue where program hangs when outputting ICP file Update & fix PT Timetable to consider boarding penalties in

calculating transit bundle top leg costs and set appropriate cost spread in route enumeration.

Cube Voyager Release Summary 3

Version 6.4.3

Fix for LINKO SKIP0=T producing incorrect set of links in the output file

Fix to test for line-line legs to ensure both start node/alighting point from transit line and end node/boarding point to transit line are present.

• Highway Fix for issue where the HWYLOAD.DLL is not loading

properly Support for networks with very large numbers of screen

lines for ICP and screen line outputs Updated memory allocation routines Added error check when using ESTMICPO FORMAT=0 and

DISTRIB_ESTM=1 Fix for voyager crash issue when FILEI TOLLMATI and MW/

ESTMO is used


Version 6.4.2

Version 6.4.2• Added support for OMX-Cube matrix conversion

• Fix for crash due to foreign characters in file names

• Fix for Build Network from Shape File function not generating output network

• Matrix Fix for ZDATI space/tab delimited data in MATRIX program Fix for BSEARCH crash

• Network Increased limit for attribute names in EXCLUDE statement

• Public Transport Additional PT crowding statistics Added keywords to control distance and time decimal

precision in PT LINKO output Fix for multiple screenline output issue Fix for out of memory issue when writing large PT data to

geodatabase Fix for PT route enumeration crash due to array boundary

violation Fix for PT crash when reading in large non-transit leg file Add check for EFARE parameter in PT when considering

fare in best path enumeration

• Highway Added support for toll gate to toll gate matrix output Added support for U-turns in highway path building Added support for U-turns in intersection modeling Added support for free turns modeling in signal junctions Support for PATHO in highway intrastep cluster


Version 6.4.2

Support for SUBAREAMATO in highway intrastep cluster Added intersection queue data output from highway

assignment Fix for delay calculation with bus blockage Fix for highway crash with MDB input Fix for highway PATHLOAD statements with matrix names

for input trips


Version 6.4.1

Version 6.4.1• Fixed an issue with print file locking by Voyager.

• Fixed bug in REPLACESTR and REPLACESTRIC functions.

• Utilities Fixed TPP2UB (Voyager Matrix to US FTA User Benefits

Matrix) crash issue.

• Highway Updated the Lambda calculation methodology for better

convergence.

• Matrix Fixed bug related to matrix RENUMBER statement. Fixed issue with reading CSV file with long fields. Fixed incorrect number of field names for geo-database

tables. Fixed issue when sorting more than five fields.

• Network Fixed crash issue when using empty lookup file. Fixed bug with NETO output to geo-databases. Fixed bug where PRINT FILE= statements overwrites files

instead of appending.

• Public Transport (PT) New FORMAT keywords to output text and CSV files for

STOP2STOP data.


Version 6.4.0

Version 6.4.0• Voyager 6.4.0 is built on 64-bit code. This provides more

stability to Voyager program runs and also provides Voyager access to all of the system memory without any restrictions.

• Voyager will not support 'if then else' statements. These statements should be modified to 'if else'. The usage of 'then' is not supported anymore.

• New/Revised Functions

(See General Syntax > Control Statement Syntax > Expressions > Numeric functions in the Cube Voyager Reference Guide): CurrentTime() - returns the local time in number of seconds

since 1/1/1970, precision down to milliseconds. FormatDateTime(datetimevalue, width, dec, formatstring) -

formats a datetime value (as returned from CurrentTime) according to the format string.

Format (datevalue, width, dec, formatstring) - this existing function has been revised slightly. It's handling of m and d in the formatstring input matches FormatDateTime above.

Better identification of errors when there is an error inside an IF statement. When there is an error in IF statement, it does not produce an error but corrupts the IF/ENDIF count. In a big script with many IF statements, it is very difficult to find the problem. This fix will identify the error statement and not point to the IF/ENDIF mismatch.

Voyager's parsing logic updated to purge any double backslashes resulting from prepending or appending directories, in particular the root directory. This will prevent issues with directory paths coded in PILOT steps in some cases.

• Highway Fix for bugs in ESTM process.


Version 6.4.0

• Reset the record counter for each ESTMICPO file, previously it continues to cumulate and reported incorrect totals.

• Change the "no routes intercepting link x to y of screen line z" warning process to test for any fractional number > 0. Previously it ignores volumes less than 1 and reports no routes intercepting link even if there are small (< 1) volumes on the screen line.

Improvements to speed up ESTM processing. Added keywords to distribute the ESTM processing.

• DISTRIB_ESTM - keyword to control if ESTM processing is distributed.

• ESTMPATH - Specifies the file path for saving the temporary ESTM files when doing intra distribution.

DISTRIBPATH - Specifies the temp file folder for an intra-distribution run.

USE_V_ONLY - Parameter Option to stop computing all the individual volume sets, when in the adjust phase, to speed up the run when individual volume sets are NOT used in the TC and COST functions.

Update ESTMS to correctly output number of ICP values in file header record.

Fix for HIGHWAY process hanging (looping continuously) in the ADJUST phase when modeling All way stop sign intersection. Only happens in very rare situations.

Fix for bug generating ICP file when running with Cluster. Change the ICP temporary file read error message

(message 762) from a warning to an error so it will stop the run instead of continuing and produce a bad ICP file.

The TIME and COST variables are now held as double-precision floating-point, enabling path building to use double-precision calculations. Assignments with the Frank-Wolfe algorithm may differ slightly.


Version 6.4.0

Added keywords in PATHLOAD statement to provide users with more control and additional capability to the highway path building process.

• MAXPATHCOST - restricts the path builder to build path only to zones/nodes that are <= MaxPathCost except when MaxPathCost = 0, which means no restriction.

• FromNode - build paths starting from list of nodes.

• ToNode - build path ending at list of nodes. New global parameter TRACESUMMARY which controls the

path trace printing to the print file. The default will be the same as before, printing the whole path. If it is set to T, then only the last link will be printed. In that case you will most likely not print A, B and any other link attributes unless you want to know which link is the last link in the path. If you print I, J and _PathCost, you will get the total cost of the path.

Additional convergence criteria (AllowLambda0).

• AllowLambda0=T/F - Previously Lambda is averaged between the high and low value of the increment it falls into, making it impossible for Lambda to reach zero (averaging between 0 and a very small number will always be > 0). This option will allow Lambda to go to zero if the low value of the increment is zero.

Added functions (CheckName, GetValue) to check for link attribute names and retrieve values. The addition of these new functions provides the user with the freedom to write scripts which are not dependent on fixed link attribute names.

• CheckName(Varname) - Check if a variable exists. If the variable does not exist, the function returns a zero. If the variable exists, the return code is a 2 digit number


Version 6.4.0

• GetValue(Varname {, DefaultValue}) - Get a numeric value from a variable or array, with or without a default value. This is limited to single variable and arrays with a default index. It will NOT work with user defined arrays or to access non-default indexed cells in an array.

Updates to junction modeling. Junction modeling now supports HCM2010 methodology for All-way stop, Two-way stop and Roundabouts. New keywords added to support the HCM2010 methodology.

• NoDefactoUnoppose - stop the default exclusive turn lane check on the unopposed turning movement.

• NoDefactoOpposed - top the default exclusive turn lane check on the opposed turning movement

• HCMVersion- specifies the HCM version to use, it can be 2000 or 2010, default value of 0 means 2000.

• SixLaneMajor=T/F - used in Two-Way stop method in HCM 2010.

• HeavyVehicles (Approach) - number of heavy vehicles, used in All-Way stop method.

• UTurns - the number of U-turns, used in Roundabout method in HCM 2010.

• CirculatingLanes (1-2) - number of circulating lanes, used in Roundabout method in HCM 2010.

• BypassType (0-2) - bypass type, used in Roundabout method in HCM 2010.

• CenterLanePercent (0-100) - center lane usage percent, used in Roundabout method in HCM 2010.

• CriticalGap - changed from a single value to 3 values, use in Two-Way stop method With HCMVersion set to 0 or 2000 and CompatibleVer set to 602, the calculations should be exactly the same as version 6.0.2.

Fix for program crash when outputting matrix file with 255 matrix tables.


Version 6.4.0

Fix for crash when PATHLOAD TOLLMATI does not use all the toll sets defined in TOLLMATI TOLLS. Additional checking has been added to stop program run when there are inconsistencies between TOLLMATI TOLLS and PATHLOAD TOLLMATI.

Added messages to check if link capacity, speed or T0 have possible issue that may affect highway assignment performance, and remind/warn users to check their network inputs.

The support for SMOOTH keyword in Bi- Conjugate Frank Wolfe assignment has been deprecated.

• Matrix The transpose operation (MI.n.n.T) may now be performed

on non-binary (e.g., DBF) inputs. MATRIX FILLMW keyword can now handle transpose

matrices. This allows the user to fill multiple temporary matrices with transposes in one statement.

New functions (CheckName, GetMatrixRow) to check for matrix table names and retrieve row values. The addition of these new functions provides the user with the freedom to write scripts which are not dependent on fixed matrix table names.

• CheckName(Varname) - Check if a variable exists. If the variable does not exist, the function returns a zero. If the variable exists, the return code is a 2 digit number.

• GetMatrixRow(Varname , MW# {, DefaultValue}) - Load a MI matrix row into a MW, with or without a default value. Varname must be a MI.x.x name.

Fix for bug when reading CSV files. The problem happens when the last line in the input file does not have CR/LF character, in that case the last character in the line is ignored, putting the wrong value in the last field of the last record, dropping the last digit or last character.

Fix for bug on RECO Form not reset to 10.2 when another RECO[] file is used


Version 6.4.0

Fix Matrix problem when using an empty DBF file in RECI. Fix for issue where FREQUENCY RANGE inconsistently

assigns values less than -1000. MATRIX program does not read-in or output EMME2 data

bank files. Fix for a memory allocation issue when using AutoMDArray. TRAM keyword usage has been updated. The keyword

value specification has been modified to Mega-Bytes (MB).

• Network The PARAMETERS NODE keyword now allows values up to

the license node limit. Fix for inconsistent CROSSTAB output. Fix for incorrect data output in network field. Fix for access violation when exporting link and node

records from NETWORK. FILEI GEOMI AFIELD/BFIELD/SEQFIELD - Additional keys to

specify the A, B and sequence fields in the input GEOMI file. Added network topology keywords/functions to analyze

network geometry. These keywords enable the user to do network geometry analysis such as link angles/movements, number of links connected to a node etc.

• SORTANGLE - Option to allow sorting the relative angle of the legs with respect to the current leg.

• SHAPEANGLE - Use shape to determine link angles.

• _N.MaxConnect - Get maximum number of connections for the whole network.

• _N.Connections - Get number of connecting nodes to N.

• _N.Legs - Get number of legs at N.

• _N.Inbounds - Get number of inbound links to N.

• _N.Outbounds - Get number of outbound links to N.


Version 6.4.0

• _NI.{name}[#] - Get inbound link attribute's value, {name} is the attribute name, [#] is the leg number.

• _NO.{name}[#] - Get outbound link attribute's value, {name} is the attribute name, [#] is the leg number.

• _NI._Angle[#] - Get inbound link angle in degrees, # is the leg number

• _NO._Angle[#] - Get outbound link angle in degrees, # is the leg number

• _NI._R_ANGLE[#] - Get inbound link angle relative to the current leg, in degrees.

• _NO._R_ANGLE[#] - Get outbound link angle relative to the current leg, in degrees.

• _L.S_Angle - Get link angle for the current link, at the start of the link

• _L.E_Angle - Get link angle for the current link, at the end of the link.

• _A.Connections/_B.Connections - Get number of connecting nodes to A/B

• _A.Legs/_B.Legs - Get number of legs at A/B

• _A.Inbounds/_B.Inbounds - Get number of inbound links at A/B

• _A.Outbounds/_B.Outbounds - Get number of outbound links at A/B

• _A.CurrentLeg/_B.CurrentLeg - Get the leg number of the current link at A/B

• _AI.{name}[#]/_BI.{name}[#] - Get inbound link attribute's value, {name} is the attribute name, [#] is the leg number at A/B

• _AO.{name}[#]/_BO.{name}[#] - Get outbound link attribute's value, {name} is the attribute name, [#] is the leg number at A/B

• _AI._Angle[#]/_BI._Angle[#] - Get inbound link angle in degrees, [#] is the leg number at A/B


Version 6.4.0

• _AO._Angle[#]/_BO._Angle[#] - Get outbound link angle in degrees, [#] is the leg number at A/B

• _AI._R_ANGLE[#]/_BI._R_ANGLE[#] - Get inbound link angle relative to the current link.

• _AO._R_ANGLE[#]/_BO._R_ANGLE[#] - Get outbound link angle relative to the current link, in degrees.

• _AI._MOVE[#]/_BI._MOVE - Get the movement number of an inbound leg relative to the current leg.

• _AO._MOVE[#]/_BO._MOVE - Get the movement number of an outbound leg relative to the current leg.

• _A.MAXMOVELEGS/ _B.MAXMOVELEGS - Variable which contains the highest number of legs per movement, in both inbound and outbound direction at the A/B node.

New functions (CheckName, GetValue) to check for link attribute names and retrieve values. The addition of these new functions provides the user with the freedom to write scripts which are not dependent on fixed link attribute names

• CheckName(Varname) - Check if a variable exists. If the variable does not exist, the function returns a zero. If the variable exists, the return code is a 2 digit number

• GetValue(Varname {, DefaultValue}) - Get a numeric value from a variable or array, with or without a default value. This is limited to single variable and arrays with a default index. It will NOT work with user defined arrays or to access non-default indexed cells in an array.

New options for CSV formats for LINKO and NODEO outputs. The new formats are:

• CS1 - comma separated values with one header line with a list of the field names

• CS2 - comma separated values with two header lines, first line has 2 values: number of fields and number of records. The second line contains a list of field names, same as line 1 with CS1 format.


Version 6.4.0

• Public Transport (PT) Modify PT process cost calculation in time tabling. New keyword DEC in stop to stop output analysis to control

the decimal precision of passenger volume outputs. Update to the ACCUMULATE=ADJACENT output for stop to

stop analysis. This setting will now additionally output mode information for each movement.

Updated PT Time Tabling route enumeration process. So the multi-routing process will always include minimum cost path.

Fix for PT crash when saving large LINKO files. Fix for a memory allocation problem which causes crash of

a PT crowding run with certain data Fixes for issue where PT Time Tabling does not enumerate

all expected routes, cannot find path with QUICKESETPATH=F, and SPREADCONST.

Fix for missing SHORTNAME and STARTTIMES in LINE output.

Fix for issues related for fare calculation. Fix for inconsistent results between FREQUENCYHDW vs

FREQUENCY. Fix for a PT fare issue when using fare on BESTPATHONLY

enumeration. Fix for PT time tabling report file where wrong departure

and arrival time values are being printed. Update to PT route enumeration to include transfer

penalties when PT time tabling is used and QUICKESTPATH=T.

Modify PT fare calculation to include the 64-bit higher precision variables.

FILEO LINKO now supports additional file formats (CSV and TXT).


Version 6.1.1

Version 6.1.1This section lists new features found in Cube Voyager Version 6.1.1.

• Cube 6.1.1 supports ArcGIS version10.2.2


Version 6.1.0 Service Pack 1

Version 6.1.0 Service Pack 1This section lists new features found in Cube Voyager Version 6.1.0, Service Pack 1:

• Fix in Matrix, in the reporting of totals in Trips/MINUTP matrices. In the Print file, the intrazonal totals are now correct.

• In Public Transport, SELECTLINK(MODE=...) functionality problems involving lists of modes have been fixed to return all of the correct modes. Previously only the first mode in the list was processed correctly. A problem where a mode range (MODE=x-y) could also result in a following expression returning a zero matrix has also been fixed.

• Cluster A problem where Highway Cluster processing of matrices

resulted in incorrect matrix summation results (in the print file) has been fixed.

Enhancement to WAIT4FILES variable substitution processing to remove any leading or trailing spaces for each file in a list of files to be waited upon

• Changes for Analyst Drive are documented in the Analyst Drive Reference Guide (available on the Cube help menu).


Version 6.1.0

Version 6.1.0This section lists new features found in Cube Voyager Version 6.1

• Support for Windows 8

• Cube 6.1 includes the new Analyst Drive.

Analyst Drive is a program using state-of-the-art data assimilation techniques to optimize trip origin-destination (OD) matrices for a given traffic assignment and input traffic count data. While Analyst Drive builds on the estimation tools provided to the Cube software suite by the Analyst module, it is a completely new program with advanced capabilities not previously available to the Cube platform.

For more information, please see the Analyst Drive Reference Guide.

• Application Manager and Cube Voyager now feature full support for filenames and paths containing foreign characters.

For more information, see Getting Started > Language settings in the Cube Base Reference Guide.

• Highway Highway, Avenue and related compontents (junction

processing, paths) now support 1000 Volume Sets, increased from 20.

Fix for the calculation of FACTOR, applicable to Conjugate Frank-Wolfe process with equation-based Lambda search. During convergence, FACTOR now correctly sums to 1 over all iterations.

Implementation of a floating point stability fix for path building

Changes to the Bi-Conjugate Frank-Wolfe process:

• A fix for computation error of the FACTORs when using EQUATION-based Lambda search in the Bi-Conjugate Frank-Wolfe process.


RG_CubeBase.pdf

RG_AnalystDrive.pdf

RG_AnalystDrive.pdf

Version 6.1.0

• There is also modification to smooth the tail of the convergence curve, and reduce the minor differences between cluster and non-cluster runs.

To support this change, Highway now features the SMOOTH parameter. This specifies the number of iterations to be used in the smoothing process of the Bi-conjugate Frank-Wolfe algorithm. To access this, add the SMOOTH keyword immediately after the COMBINE=EQUI ENHANCE=2 statement. The valid range of this keyword is 0 - 99. SMOOTH defaults to 10 when unspecified.

For example, if SMOOTH=x, then the smoothing process will be executed every x iterations. Note that, SMOOTH=0 means no smoothing process will be executed.

• Finally, backwards compatibility with Bi-Conjugate Frank Wolfe from Cube 6.0.2, if desired, can be enabled using the new BiFWVERSION subkeyword. Not compatible with SMOOTH.

For example: COMBINE=EQUI ENHANCE=2 BiFWVERSION=602

Fix for issue where Subarea extraction would generate small amounts of data within non-existing zones

Fix for TURNVOLO outputs, when VARS= is selected. Previously, incorrect format parameters and volumes were selected if volumes were not consecutive starting from 1 in PATHLOAD/DYNAMICLOAD statements.

• Avenue Avenue, like Highway, now supports up to 1000 volume

sets. Avenue features a new PARAMETERS PRESERVEMW

keyword. PRESERVEMW will save the contents of the designated MW array (or arrays) across time segments within an iteration. This keyword is |KIVPa|.


Version 6.1.0

The calculation method of RelGap now matches that in Highway. Link cost values are also now recalculated for each time segment.

Fixed Avenue path skimming error for TRACE expressions without __TS__ expansion. Whenever the offset (time parameter) exceeds the first time segment, the skim is now correct.

Fix for TURNVOLO outputs. Previously, incorrect volume array elements were selected when volumes were not consecutive starting from 1 in PATHLOAD/DYNAMICLOAD statements.

Additional stability improvements

• Bugfix in Junction keyword processing. When an intersection was converted to Empirical Roundabout using Intersection Data Editor, leftover keywords from the previous method could cause other required keywords for the Empirical Roundabout to be ignored. This fix ensures that the required keywords are not ignored.

• Matrix Matrix may now read, process and generate files of any size,

subject to available system memory and storage. Matrix now supports the DBI.#.JOINCOND variable. This is

an integer indicating the status from a JOINTODBI action, i.e. when the "Join To" DBI file is moved to a different record. The outputs are:

• 0 = a match was found

• 1 = error (not common)

• 2 = match not found and the current record is the next higher key record

• 3 = match not found and the search for key is greater than the last record; the current record is the last record

• -1 = no match found within in the chain of joins; the current record is not changed


Version 6.1.0

• Public Transport now features the optional PARAMETERS MODIRE keyword, enabling a modified route enumeration process. Use MODIRE to enhance the stability of the enumerated route sets and skims under small variations of network characteristics.

• In Cube Cluster, you may represent Wait4Files Files as a comma-separated, non-spaced list of files in a Pilot variable. For example:

MYFILES=’TestDist1.script.end,TestDist2.script.end,TestDist3.script.end’

Wait4Files Files=@MYFILES@, CheckReturnCode=T,UpdateVars=vname,Matrix.xname, PrintFiles=Merge, DelDistribFiles=T

• TRNBUILD’s node number limit (from the input highway network) has been increased from 999,999 to 10,000,000.

• Build Network from Feature Class, which is launched from Voyager script, has refinements to the Consolidate AB/BA Field Pair function — and its application of field masks. For more information, see Utilities > Build Network from Feature Class in the Cube Voyager Reference Guide.

• The path building functions, available from the command line via Voyager.exe, feature a new PRINTMODE=SUMMARY parameter. This outputs cost information for each path without detailed data on each node, speeding up the process and producing smaller path files. For more information, see “Shortest Path” in the Utilities

Chapter of the Voyager Reference Guide.

• Voyager now supports the hiding of script contents from the output print file. This may be done by: Using the Hide Script check box in the Cube Voyager

window, or Using the /HIDESCRIPT command line option with

Voyager.exe

• Documentation — see Cube Voyager Reference Guide.


RG_CubeVoyager.pdf

RG_CubeVoyager.pdf

RG_CubeVoyager.pdf

RG_CubeVoyager.pdf

Version 6.1.0

There is a new section describing the Probit and Burrell stochastic traffic assignment methods. Go to Highway Program > Theory > Stochastic Traffic Assignment.


Version 6.0.2

Version 6.0.2• Cube Voyager:

Voyager.exe has been completely rewritten, with improved speed, stability, and reliability with large cluster runs.

• Cube Cargo: The ‘FINE DISTRIBUTION’ and related Cargo errors under

Windows 7 and Vista are now resolved.

• Pilot: PARAMETERS now supports the VARDECIMAL keyword. It

specifies the decimal place precision for numeric variables passed between Pilot and individual program modules, using the LOG statement and @...@ tokens . This will allow an user to increase the precision of the passed values as needed.

This value also controls the decimal precision of variable values logged in each program module so that small values can be passed back to Pilot.

Default is 5. Minimum is 0, maximum 15.

• Highway: The Voyager Reference Guide now includes a detailed

section on multiplie user class assignment, including an example using FUNCTION COST. See Example 8 in the Highway Chapter of the Cube Voyager Reference Guide.

• Network: FILEI now supports the GEOMI[x] keyword. These are link

input files to Network to act as geometry sources for output GDB networks and shape files. As before, the source of the geometry is specified in the LINKMERGE phase by setting the GEOMETRYSOURCE variable.

FILEO LINKO/NODEO FORMAT now include the SHP option. This will write a shape file with geometry from the GEOMI or LINKI geometry source(s) but the associated DBF file will be


RG_CubeVoyager.pdf

RG_CubeVoyager.pdf

Version 6.0.2

from the LINKO/NODEO process. This also allows using the INCLUDE and EXCLUDE keywords to limit the fields written to the shape file DBF.

For more on these Geometry features, and an example, please see the Network chapter of the Cube Voyager Reference Guide.

NETWORK FILEI NODEI can now read a standalone node shapefile as an input. The shapefile must have an associated DBF file. NETWORK will create a new DBF with all fields and data as the original shapefile. It will also add X&Y fields to contain the associated point coordinates.

• The size limit of a single print field in a PRINT statement is increased from 5,000 to 50,000. You can now copy a long record up to 50,000 characters from RECI to PRINTO in a single print field. See Matrix changes, below.

• Matrix: The RECI record buffer is increased from 10,000 characters

to 32,000 characters. If a record is over 32,000 characters, MATRIX will issue an error message.

The limit on PARAMETERS MAXSTRING is increased from 3,000 to 32,000 characters.

The RECI FIELDS repeat count is increased from 1,000 to 16,384, allowing definition of 16,384 fields.

RECI FIELDS now works correctly for specification such as FIELDS=3,7,9 has been corrected.

Matrix has a new sub-keyword SKIPRECS in MATI. This applies to text inputs, allowing the script to skip over a number of records before processing the data records.

ARRAY control, TYPE keyword: maximum length of strings in a string array is increased to 999

FILEI control, RECI keyword: there is now more descriptive output of record processing output in task monitor. For example:

Reci: 30000 of 113000 (26%)


Version 6.0.2

• Cluster: New STARTHIDE command line argument. Use this to

minimize cluster nodes to the system tray and reduce window clutter. There is also now a ‘Hide New Nodes’ check box in the Cluster Node Management dialog box performing the same function.

• Public Transportation: Keywords of PT service model are updated. The wait curves

are adopted in the Service-frequency model and Service-frequency-cost model to better approximate the wait time when identifying the basic sets for transit choice. Please see CUBE VOYAGER > PUBLIC TRANSPORT PROGRAM > CONTROL STATEMENTS > FACTORS > SERVICEMODEL for detailed description.

Note that, the computation results will be changed with the updates. To keep the results unchanged, the keywords FREQUENCY and FEQUENCYCOST in the script must be replaced with FREQUENCYHWD and FREQUENCYCOSTHWD respectively.

Fix for SHORTESTWALK factor, where the behavior of the T & F cases was incorrectly switched.

Fixes and enhancements to Timetabling model

• PT now supports the FACTORS keywords:

• QUICKESTPATH — When true, enumerates a single path. Default is false.

• QUICKESTMULTI — When true, enumerates a single fastest path for each "departure time." Default is false.

• PT timetable consistently considers the PERIOD start time as the earliest departure time at origin

• Updated logit model in Timetabling when BESTPATHS=T and KIRCHOFF=F, i.e. using logit model to calculate probability for each path (best path for each departure time), rather than setting crude avearge


Version 6.0.2

for each one (which gives same probability for each path). PT will also print a message to report the model adopted in route evaluation in this case.

• The new Cube Base Path Building functions may be run from the command line, so they can be called from Voyager. For more, see: Version 6.0.2 in the Cube Base Release Summary. “Shortest Path” in the Utilities Chapter of the Voyager

Reference Guide

• The Cube Voyager Reference Guide has new content: There are new sections on running, from Voyager scripts

and/or the command line:

• Link Consolidation

• Calculating Shortest Path (see above)

• Building Networks from Feature Class (Shape).

The new sections are in the Utilities chapter. The General Syntax chapter has a new section describing

Voyager Return (Error) Codes.


RG_CubeVoyager.pdf

RG_CubeVoyager.pdf

RS_CubeBase.pdf

Version 6.0.1

Version 6.0.1This section lists new Cube Voyager features found in Cube Voyager version 6.0.1.

• In Highway FILEO:

ESTMDATO output screenline files, and ESTMPICO output intercept files, may now be indexed up to 20 files.

ESTMDATO now includes LINKTOICP subkeyword, specifying which ICP file the output is linked to.

ESTMPICO now includes the FORMAT subkeyword, supporting a selection of Analyst output formats.

• In Highway PATHLOAD:

ESTMO is now an integer, supporting the FILEO changes described above.

• Fixes to Cluster in the management of print files, and matrix files with small number of zones per slave

• Public Transport has new GENERATE keywords supporting the AUTOCON methodolgy, for developing drive access links for the transit networks:

CONNREPORT DISTMAT PREMIUMMODE

GENREPORT INTERNALZONES MODEPRIORITY

PERIOD CBDZONE INFLTRANSITFARE

TIMEMAT DISTANCEFACTOR INFLAOC

VOT OVTRATIO AATRATIO

CHECKRELEVANCE CHECKBACKTRACT MAXBACKDIST

AUTOCCPNR AUTOCCKNR AOC

MAXCONN MAXCONN MAXCONN

AUTOMODE AUTOMATCH PNR


Version 6.0.1

By using these new keywords in the PT GENERATE control statement, the following functionalities are realized:

1. Generation of Park and ride (PNR) and Kiss-and Ride (KNR) drive-access connectors, where the mode number for PNR, KNR and auto access can be specified by the user with keywords.

2. Skipping generation of connectors for certain user defined zones.

3. Overriding of specific connections, either by deleting or replacing them.

4. Calculating weighted costs for the connectors. This weighting takes into account:

• Value of time

• Inflation on various cost components (transit, parking etc.)

• Out-of-vehicle time versus transit-in-vehicle time ratio

• Drive access time versus transit-in-vehicle time ratio

• Auto occupancy for PNR and KNR

5. Penalizing drive times that are outside defined distances from the stations.

6. Connectors are checked for "relevancy" (i.e. It will not generate a connector to a station farther away but in the same general direction of a closer station).

7. Connectors are checked for "backtracking" (i.e. When a station is on the other side of the CBD or a station is in the opposite direction of the CBD).

8. Eliminating connections to stations if the CBD is closer and the station does not have a premium service.

KNR XTRAAUTONTLEGI SKIPZONES

INFLPARKINGCOST ORIGINTERMTIME MAXBACKFACTOR

MAXCHECK KNRMODE


Version 6.0.1

9. If the second shortest connector is ignored because it is in the same general direction of the first shortest connector and CHECKRELEVANCE is True, the program considers additional connectors farther away for other "relevance" connectors.

10. More connectors are considered to identify reasonable connectors if the top two connectors are over the backtracking limit. The maximum connectors to consider and generate can be set by the user using two parameters.

11. Stations with no PT service are ignored instead of being lumped together in one set.

• Additional fixes and enhancements.


Version 6.0.0

Version 6.0.0This section lists new Cube Voyager features found in Cube Voyager version 6.0.0:

• Public Transport: New BESTPATHONLY subkeywords:

• EVALFARE — will enable all the fare evaluation features available to the full PT multi-routing analysis. This includes support for all of the fare systems supported by Voyager's PT and all of the FAREA and FAREP options available in the SKIMIJ phase. However, this option will still not impact the optimal route identified by the BestPathOnly route enumeration.

• ENUMFARE — will enable a simplified fare model during the BestPathOnly route enumeration. In this case, the optimal route identified will have taken into consideration this simplified fare model during the path building. The user may specify one of two fare systems, DISTANCE or FLAT. A new parameter EFARE tells the program that the descriptions of this ENUMFARE system are available in the FAREI file.

New PARAMETERS keyword:

• EFARE — see ENUMFARE above. New PARAMETERS HDWAYPERIOD subkeywords:

The HEADWAY[p] (and HEADWAY_R[p]) line keywords are indexed by time period so that the users may code several values in one dataset that can be selected dynamically based on the HDWAYPERIOD parameter in the script.

In order to provide better ways to calibrate and adjust transit travel times, some of the node specific keywords (specifically DWELL and DELAY) have also been made available as a line level keyword (specifically DWELL_DEFAULT and DELAY_DEFAULT) which can provide a default value for the line which is then able to be overwritten with the node keywords.


Version 6.0.0

The line keywords are indexed by period as well. In this way, several defaults may be stored and called dynamically. The following keywords have been updated to use the HDWAYPERIOD parameter to set the value dynamically:

• DWELL_DEFAULT

• DELAY_DEFAULT

• TIMEFAC

• Numerous other bug fixes and enhancements


Version 5.1.4

Version 5.1.4This section lists new Cube Voyager features found in Cube Voyager 5.1.4:

• Junction modeling: (5.1.3) Improvements to single-shared-lane-approach

modeling for Signal, TWSC, and Priority junction models. The OPPOSED movements are modeled separately from the THROUGH / UNOPPOSED movements.

Improvements to the major approaches modeling for Two-way Yield / Priority junction model

Several other bugs fixed

• Cluster Intrastep support introduces flexible wait for slave node

readiness. It waits up to 10 seconds, but finishes earlier if all .script.cmdend files are ready for each node specified.

Improved support for scheduling a MULTISTEP job to the same MULTISTEP node executing the script

• Public Transport Corrections to syntax of XFERPEN, XFERCONST and

XFERFAC. Previously, values assigned by earlier statements could be overwritten by those defined in later statements

• Avenue Fixed static data initialization issues preventing multiple

runs with different sized networks Fixed temporary path file issue where it was not being

closed after Avenue run

• TRNBUILD Improved support for circular PT lines that have nodes that

are exit only (ACCESS=2)


Version 5.1.3

Version 5.1.3This section lists new features found in Cube Voyager version 5.1.3:

• Licensing algorithms expanded to allow two simultaneous processes with a single license.

• Junction modeling: Improvements to Progression Adjustment Factor when

doing junction modeling Improvements to single-shared-lane-approach modeling

for Signal, TWSC, and Priority junction models. The OPPOSED movements are modeled separately from the THROUGH / UNOPPOSED movements.

• Avenue Fixes for a bug in the capacity calculation. Changes to provide better support for sequentially running

multiple scenarios.

• Expanded PT Editing capability for the Time Tabling keywords DAYTYPE, STARTTIMES, AT, ARR, and DEP

• Added the capability to use expressions in MW index on the left side of equations. Previously, only arrays could have an expression index on the left side. For example, the following is allowed now: MW[x+5][y*2]=....

• PT Crowd Modeling now outputs an intercept file (*.icp)

• New content in the Cube Voyager Reference Guide, "Intersection Modeling" chapter, describing theory and reference material for various intersection control types

• Numerous other bug fixes and enhancements


Version 5.1.2

Version 5.1.2This section lists new Cube Voyager features found in Cube Voyager version 5.1.2:

• Numerous bug fixes and enhancements

• Field names In DBF files can now contain 12 characters instead of 10

• Clarifications and updates to the PT documentation (see Cube Voyager Reference Guide).


RG_CubeVoyager.pdf

RG_CubeVoyager.pdf

Version 5.1.0

Version 5.1.0This section lists new features found in Cube Voyager Voyager version 5.1.0:

General

• Added ability to detect multipoint feature class and convert it to a single point feature class that can be used with Voyager.

• Stand alone comments using either ; or /* */ will no longer be copied to report files.

• Inclusion of new statistical distribution functions; ExpINV, ExpDIST, GammaINV, GammaDIST, LogNormInv, LogNormDist, PoissonINV, PoissonDIST, NormalINV, and NormalDIST.

Matrix Program

• Added support for multi-dimensional arrays in the Matrix module. Note: Syntax changes have been made affecting previously coded character arrays

Highway Program

• Added new Conjugate and Bi-Conjugate Frank-Wolf Highway Assignments.

• Added new Stochastic Highway Assignment Algorithms.

• Added new Path-Based Gradient Projection Highway Assignment Algorithm.

• Added dbf turn volume file handling to path building.

Public Transport Program

• Added the ability to incorporate public transport schedules with the PT Modules.


Version 5.1.0

Avenue Program

• Added new packet allocation methods.

• Added the ability to incrementally simulate time segments.

Land Program

• Multiple submarkets in real estate market

• General, user-defined bid functions

• New general, user-defined location externality functions

• New general, user-defined rent function

• New general, user-defined cost function

• General linear restrictions on location and supply


Version 5.0.2

Version 5.0.2Version 5.0.2 enhances syntax error checking in scripts. For example, Cube Voyager now checks that each ENDIF statement has a corresponding IF statement. As a result, scripts containing syntax errors and that ran successfully in previous versions might fail in this version.


Version 5.0

Version 5.0 Version 5.0 enhances Cube Voyager to read from and write to the geodatabase in model applications. This section lists features introduced in version 5.0 of Cube Voyager.

General

• Includes all cumulative minor bug fixes since the release of version 4.2.

Highway program

• Adds support for storing and maintaining highway network data in the geodatabase file format. You can link input and output networks from the Highway program directly to a network feature class in a geodatabase.

Network program

• Adds support for storing and maintaining highway network data in the geodatabase file format. You can link input and output networks from the Network program directly to a network feature class in a geodatabase. You can also link input and output link and node data directly to feature data classes in a geodatabase.

Matrix program

• Adds support for storing and maintaining data tables in the geodatabase file format. Specifically, the Matrix program can read geodatabase data tables with the FILEI DBI or FILEI RECI control statements. The Matrix program can write to geodatabase data tables with the FILEO RECO control statement.


Version 5.0

• Adds support for reading from and writing to EMME/2 data banks. The Matrix program can now write matrices, zonal data, and constants to an existing EMME/2 data bank, and the Matrix program can read existing data in an EMME/2 data bank and save in Cube Voyager binary or DBF formats.

Public Transport program

• Adds support for storing and maintaining transit data in the geodatabase file format. You can directly link input and output transit line, link, and node data from the Public Transport program to line, link, and node special feature classes in a geodatabase.

Cube Avenue program

• Optimization improvements to the ILOOP phase improve run-time performance.


Version 4.2

Version 4.2This section lists features introduced in version 4.2 of Cube Voyager. Categories include:

• General

• Highway program

• Matrix program

• Public Transport program

• Cube Avenue program

General

• Includes all cumulative minor bug fixes since the release of version 4.1.2.

• Modifies Cube Voyager modules to support FILEI/FILEO directly to and from a geodatabase file format. Starting in version 4.2, you can link networks and data tables directly to their feature class in a geodatabase.

• Adds several new character functions for use in expressions.

Highway program

• Adds a new path-based toll capability that allows point-to-point tolls to influence path building. This is a major enhancement for modeling toll road systems.

The PATHLOAD statement has a new keyword, TOLLMATI, which you use to specify the input toll matrix file that contains the toll system’s on-off tolls. For example, FILEI TOLLMATI[#]=.

• Adds a new automatic link consolidation process as an option in path building.

Using optional keyword, CONSOLIDATE=T, on the PATHLOAD statement consolidates links that are part of the same road segments prior to path building. Consolidating the links


Version 4.2

reduces the size of the link table and thus reduces path-building time. Depending on the input network’s efficiency, reductions in run time could be significant.

• Adds new subkeywords, COUNTVAR, CONFVAR and DEFAULTCONF, to the ESTMICPO keyword under the FILEO statement. With these subkeywords, you can define both the count variable and the confidence value variable used to create the screenline and intercept output files for Cube Analyst.

• Adds the FILLMW statement previously only available in the Matrix program to the ILOOP phase of the Highway program.

Matrix program

• Adds a new database processing capability, DBI, that allows multiple input data files.

Triggered by the keyword DBI[#]= following the FILEI statement, DBI is similar to record processing with FILEI RECI. There are three main subkeywords: SORT sorts DBI data on input. AUTOARRAY directly populates internal arrays from data

fields in the DBI data. JOINTODBI joins multiple DBI files on one or more common

fields.

Three new functions support DBI processing: DBIReadNext, DBIReadRecord, and DBISeek.

• Supports defining an ARRAY as a string array in addition to a numeric array.


• Adds the outputs Cube Analyst requires to estimate transit matrices.


Version 4.2

New keywords, INTERCEPTO and SCREENLINEO, in the FILEO statement extract the intercept files and screenlines files by user class, as required for public transport matrix estimation in Cube Analyst.

• Improves Public Transport reporting.

The LINKO and STOP2STOPO keywords on the FILEO statement now allow up to four output files. Additional subkeywords allow you to control the content of the specified files: For LINKO, subkeywords provide two additional data

structures for the output DBF file:

• ONOFF outputs boarding and alighting counts from the loading process.

• ONELINKREC outputs a file with one record per link. The record contains transit vehicles per hour, total transit loadings, and total nontransit loadings. You can output volumes by user class. For crowd-modeling runs, you can output flow-metered values.

For STOP2STOPO, you can now recode transit modes before analysis. As a result, you can treat several modes as the same mode when obtaining STOP2STOPO results.

• Increases the limit on the number of fare systems (specified with FARESYSTEM) from 999 to 1999.

• Implements an improved file format for the route file (*.RTE) when using the ROUTEO keyword on the FILEO statement. The improved file format improves run times for route evaluation.

Cube Avenue program

• Improves run-time performance and memory usage.

• Adds a new keyword, PACKETLOG, to the FILEO statement that writes a detailed packet log file. Additional subkeywords let you control the content of the file. You can open the file with Cube to view the packet animation.


Version 4.2

• Adds the MW keyword to the DYNAMICLOAD statement. You can use this keyword to dynamically skim path costs.

• Adds new parameters, CAPLINKENTRY, PKTPTHSIZ, and MAXPTHPERSEG, to give you more control.


Version 4.1


• General

• Pilot program

• Highway program

• Matrix program

• TRNBUILD (TP+)


General

• All cumulative minor bug fixes since the release of version 4.0.1 have been included in version 4.1.0.

• New optional Cube Voyager program: Avenue. Avenue implements dynamic traffic assignment capabilities into the Cube Voyager Highway program.

• New optional module Cube Cluster. Cube Cluster implemented in Cube Voyager allows you to create your own high-performance computer cluster to distribute the workload of your voyager model run across multiple computing nodes. Each computing node consist of a single computer processor so assembling your own computer cluster is as simple as networking several existing computers together on a local area network or running on a dedicated multiprocessor machine.

Pilot program

• Cube Cluster is an add on program to the Pilot program

• The operation of the *command has been modified. The *command now opens a minimized system command window to perform the requested commands.


Version 4.1

• A new **command has been added for backward compatibility. The ** command opens a system command window that is NOT minimized. This option allows the command window to stay visible on the desktop when such visibility is desired.

Highway program

• Cube Avenue is an add on module to the Highway program

• DistributeINTRASTEP added to support INTRASTEP distribution of zonal processing in the Highway program in support of Cube Cluster.

Matrix program

• DISTRIBUTEINTRASTEP added to support intrastep distribution of zonal processing in the Matrix program in support of Cube Cluster.

• XCHOICE added to replace existing CHOICE statement for implementing logit choice models. CHOICE is retained for backward compatibility but as of v4.1 XCHOICE should be used. XCHOICE was implemented significantly improve the run time of logit choice models.

TRNBUILD (TP+)

• The network size limit of 65000 links in TRNBUILD has been lifted to be compatible with the rest of the system with this release (999,999 links, 999,999 nodes, and 32,000 zones).


• Near the end of the development cycle for v4.1.0 as part of implementing the BESTPATHONLY factor, a bug was discovered in the enumeration process of the released version. This bug allow some routes to be enumerated in some situations where they should not have been. This bug was corrected in the


Version 4.1

enumeration process of v4.1. Due to this correction, results from a PT run using v4.1.0 may not match exactly the results obtained from a run using the prior released version.

• PT Select Link. A new select link function produces output matrices of demand matching selection criteria, and can also perform select-link loading. If multiple user classes are modeled, select link outputs may be produced for each class from the individual enumerated route files. The select link function provides a range of facilities to identify travel demand using particular links, lines, or nodes in the network. Lines may be identified explicitly or selected based on their mode or operator attributes.

• Loading PT demand which meets a select link criterion, for subsequent display or analysis.

• Paths may now be enumerated to match a user specified set of modes with the FACTORS MUSTUSEMODE.

• An alternative to multipathing is now available with the new FACTORS BESTPATHONLY.

• Route enumeration and evaluation based on a best-path methodology

• A facility to trace and REPORTO the details of routes

• Increase in maximum permitted number of generated connectors from any zone (or node) to 999 non-transit legs.

• The GENERATE command has been improved at release 4.1. In earlier versions the GENERATE command excluded NT legs between pairs of nodes where a transit leg ran, but it has been found that this is not always appropriate. To create exactly the same NT legs as obtained from earlier software versions, the user should set the new keyword XFERMETHOD=1. When changing to release 4.1, the newly created NT legs may cause the loss of some connectors which had previously been obtained. This is due to the MAXNTLEGS keyword which specifies the maximum number of connectors generated by mode. Where one of the newly allowed connectors has a low


Version 4.1

cost, it may replace a higher-cost old connector. To use the new method of generating transfer connectors but not loose any connectors the older process had generated, specify XFERMETHOD=2. The default value XFERMETHOD=0 will use the improved process implemented beginning with v4.1.


Version 4.0


• General

• Pilot program

• Highway program

• Network program

• Matrix program

• Generation program


General

• Beginning with v4.0, all Voyager modules have been compiled with new Microsoft compilers. Versions prior to v4.0 were compiled using Borland compilers. The primary reason is modernization. This is a change that is past due and Citilabs felt we could put if off no longer. We have found that the two compilers give very similar results, but slight difference in model results may be detected.

• The LOOKUP statement now supports the reading of data in DBF format via either the LOOKUPI or FILE keyword. Limitations on LOOKUP data size have been modified to allow an unlimited number of LOOKUP records and up to 30000 LOOKUP curves with ascii data or 255 curves with DBF data.

• New Trig functions are now supported in expressions and computations. SIN, COS, TAN, ARCSIN, ARCCOS and ARCTAN are now available.


Version 4.0

Pilot program

• New ONRUNERRORGOTO and CLEARERROR statements added to provide additional flexibility in flow control to allow conditioning on, reporting on and continuing after error returns from program modules

• New SENDMAIL statement added to allow sending mail and attachments based on model run conditions. For example, generate and send an e-mail notification when a model run completes or send an e-mail notification on specific failure modes or result conditions.

• New DOWNLOAD statement added to allow requesting file downloads from a remote location via a URL file address.

Highway program

• A new optional CONVERGE phase has been added to the module to allow the user to specify conditional convergence criteria. For instance if you want the assignment to continue until the GAP cutoff value has been achieved or exceed at least 5 times and the number of iterations is greater than 25, this can now be specified.

• A new PARAMETERS keyword, RELATIVEGAP, has been added as an additional equilibrium closure statistic.

• New PARAMETERS statements TCCOEFF and TCEXP, indexed by LINKCLASS have been added to allow coding of LINKCLASS specific coefficients and exponents for use in the TC functions. TC functions now recognize TCCOEFF and TCEXP for the specific LINKCLASS.

• A new subkeyword LAMBDASEARCH has been added following the COMBINE=EQUI keyword on the PARAMETERS statement. LAMBDASEARCH allows values of AREA or EQUATION for specifying one of two available lambda search processes during equilibrium highway assignment. LAMBDASEARCH replaces MINIMIZE=BPR/SYS.


Version 4.0

• The INCLUDE keyword on the FILEO statement has been modified to allow the user to specify and keep LW.name working variables that are created during the run on the final output network.

• The new keyword LINKIDARRAY on the PATHLOAD statement has been added to allow the identification of specific links traversed on a path. The LINKIDARRAY holds the values from a user specified LI.name or LW.name array for the links traversed on the current path. Values from LINKIDARRAY can then be accumulated into working matrices with new subkeywords LINKIDCNTMW, LINKIDMW, LINKIDLASTUSE, and LINKID#MW.

• The standard PRINT statement and its keywords have been added as subkeywords to the TRACE keyword on the PATHLOAD statement. This allows for directing the printing of path traces to an external user formatted PRINTO file instead of just to the run print file.

• New subkeywords PATHOGROUP and FULLPATH have been added to the PATHLOAD statement to allow the user more control of the specific content and size of the PATHO file.

Network program

• In v3.2.x and earlier, node and link records in a given data file had to be unique. This limitation has been lifted starting with v4.0. In v4.0, the COMBINE keyword has been added to both the LINKI and NODEI keywords under the FILEI statement so that users can specify how they would like to treat the values on the fields of redundant data records.

• New subkeyword DETAILS has been added to the LINKI keyword on the FILEI statement to support reading and accessing the iteration specific attributes from Tranplan-formatted input networks.


Version 4.0

Matrix program

• Significant changes have been made to RECI/RECO processing for v4.0 which may effect user scripts coded under earlier versions. See the documentation changes for these specific keywords and see “Working with RECI/RECO processing in v4.0 and beyond” on page 439 of the Cube Voyager Reference Guide.”

• New subkeywords SORT and DELIMITER have been added to the RECI keyword under the FILEI statement. This allows sorting input records on up to five data fields and supports user specified data field delimiters.

• New subkeywords LISTERRS, MAXERRS, MAXSCAN, and MAXRECS have been added to the RECI keyword under the FILEI statement to allow for improved error handling and reporting.

• New subkeywords FORM, CFORM, and REPORT have been added to the RECO keyword under the FILEO statement to improve formatting and reporting of RECO data.

• New keyword MATSTAT has been added to the REPORT statement to generate statistical summary reports of input, output and working matrices.

• Reading and processing of character fields with RECI/RECO processing is now fully supported.

Generation program

• New NAMES subkeyword added to the DBF keyword under FILEO PAO to allow the user to provide field names for data written in DBF format.


• Public Transport Crowding. A new facility models the higher perceived costs associated with travelling in crowded conditions (where seats are not available and standing occurs). A second enhancement models the additional wait time arising when capacity constraints prevent taking the first desired


Version 4.0

service to arrive at a boarding node. These enhancements use an iterative loading procedure to adjust transit leg costs, wait times and the probability of boarding, so redistributing demand between lines and re-routing journeys in response to crowded conditions.

• Specific modes may be excluded from runs for a user class by specification of appropriate FACTORS. This can be applied to non transit modes (for example, to restrict to walk or auto access connectors), or to exclude particular transit modes from a model.

• The route enumeration process has been improved, and is now faster. Where the minimum cost journey requires more than MAXFERS transfers the software can now enumerate this one route, so offering better connectivity between origins and destinations. This allows the use of lower values for MAXFERS (which now becomes an upper limit on multipath enumeration).

• Route files of over 2 Gb size are now supported.


Version 3.2


• General

• Highway program

• Matrix program


General

• The documentation has had minor updates and corrections with v3.2.1.

• Note that the meaning of the MAXFIELDS subkeyword on the FILEO MATO statement has been changed. In v3.2.0 and earlier, MAXFIELDS actually meant maximum sets where the sets are defined by what follows the semicolon (:) on the PATTERN definition for the output. Starting with v3.2.1, MAXFIELDS now literally means the maximum number of data fields in the pattern and not the number of sets in the pattern. If you have processes that make use of the MAXFIELDS subkeyword for matrix output you should check the format of your output and verify that it is still in an acceptable format.

• All cumulative minor bug fixes since the release of v3.1 in June 2003 have been included in v3.2.

• New keywords have been added to allow Cube’s Application Manager to provide file boxes and file control for input files associated with the LOOKUP statement and for output files associated with the PRINT statement. For all programs, the FILEI statement now has a LOOKUPI keyword and the FILEO statement has a PRINTO keyword. The LOOKUP statement now has a LOOKUPI keyword and the PRINT statement has a PRINTO keyword.


Version 3.2

• A new keyword has been added to the PRINT statement to allow direct printing of a CSV-formatted file. For all programs the PRINT statement now has a CSV keyword. Setting CSV=T will write the file in CSV format.

• The maximum number of matrices (tables) on a Cube Voyager/TP+ binary matrix file is now 255. For all programs that use binary matrix files, the maximum number of working matrices (MW) that can be defined is now 999 and is set with the PARAMETERS MAXMW.

• The maximum number of items on the LIST keyword of the PRINT statement is now 500.

• A new numerical comparison function has been added to the available expressions. CmpNumRetNum can be used to compare two numbers or expressions and return one of two numbers or expressions based on the result of the comparison.

Highway program

• The new PATHO keyword is now available on the FILEO statement to write a binary path file for display and analysis in Cube. Additional subkeywords under PATHO include COSTDEC and ITERS. COSTDEC sets the number of decimal places on the cost variable. ITERS specifies which iterations from the assignment are to be kept in the path file (the default is all iterations). In order to generate the path file the new keyword PATHO is also now on the PATHLOAD statement and must be specified. FILEO PATHO has a file name as its value and is indexed [#] where # can equal 1-10. The PATHO keyword on the PATHLOAD statement has the index number of the FILEO PATHO file where you want the paths from this PATHLOAD statement to be written. This allows the user to associate path files from multiple PATHLOAD statements to specific output files. The PATHLOAD PATHO keyword has the NAME INCLUDECOST and ALLJ subkeywords. NAME is required and provides a unique name for your path set from this PATHLOAD statement. INCLUDECOST <F> indicates if you want to include the costs


Version 3.2

associated with the paths on the path file. ALLJ <F> indicates if you want to keep paths to all J or just those J with trips assigned (default).

• The new TURNPENO keyword is now available on the FILEO statement to write a final set of turn penalties associated with both fixed turn penalties and dynamic turning delays for junction or intersection coding. When a traffic assignment is performed in Voyager the final link and intersection delays are calculated at the end of the final iteration and written to the output network and now to the TURNPENO file. In order to skim attributes from the network using these final values the Highway program must be run for one more iteration to build and skim these final paths. The new TURNPENO capability provides the final turn delays associated with the junction coding that can be included in this final skimming iteration to get final skim matrices that match the final assignment.

• When an assignment is performed the output network now includes the following additional link attributes: CSPD (congested speed), VDT (vehicle distance traveled), VHT (vehicle hours traveled).

• A new keyword has been added to the REPORT statement to turn off reporting of iteration specific turn delay reporting when performing junction based assignments. Set REPORT PENI=F to turn off the reporting.

Matrix program

• Reading record data in Matrix with the FILEI RECI keyword now supports DBF file formats.

• The new RECO keyword is now available on the FILEO statement to write output record files. The new WRITE statement is now available to direct record data to the appropriate RECO file. This feature along with RECI for DBF file types allows Matrix to function as a DBF file processor


Version 3.2


This section lists features introduced to the Public Transport program, in the following categories:

• FARES modeling

• Service frequency and cost model

• STOP-TO-STOP movements

• Nontransit network development

• Zone selection by trip matrix

• Intersection delays

• Routing algorithm

FARES modeling

A comprehensive fares modeling capability has been added. A variety of fare systems can be represented and fares can be used in route evaluation and skimming.

The new functionality is not backwardly compatible with earlier versions of Public Transport, which has been decommissioned. Fare data for versions prior to 3.2 can be converted for use in 3.2 as follows:

• Move FARETYPE control statements from the PT system file to a new file

• Change FARETYPE to FARESYSTEM

• Change FAREBASE to STRUCTURE

• Change FAREFROMFT to FAREFROMFS

• Input new file to Public Transport with FILEI FAREI

• Change FACTORS FARETYPE to FARESYSTEM


Version 3.2

Skims are activated, as before, with FAREA and FAREP. Fares may now be included in route evaluation by setting PARAMETERS FARES=T.

Service frequency and cost model

The Service Frequency & Cost Model (SFCM) has been added, which allocates trips between the attractive transit choices available at the various stages of a journey. The SFCM takes into account “time to destination” in addition to service frequency; the Service Model considers only frequency.

STOP-TO-STOP movements

A new loading analysis has been added, which extracts a variety of stop-to-stop movements for selected stops or groups of stops. The output is saved in a DBF file and, optionally, to the print file.

Nontransit network development

The following changes have been made to the GENERATE statement:

• COST, as before, is the criteria for building minimum cost non-transit legs. However, EXTRACTCOST has been added, which allows a different cost variable to be skimmed from the underlying network. The cost of the non-transit legs is EXTRACTCOST, or COST if EXTRACTCOST is not specified.

• The cost of non-transit legs can be weighted by FACTORS RUNFACTOR[NTMode]. It is, therefore, not necessary to include this weighting in COST, as in earlier versions, or EXTRACTCOST.

• MINCOST[tmodes] has been added which specifies a minimum cost for non-transit legs. Legs of less than MINCOST cost are not saved.


Version 3.2

Zone selection by trip matrix

Zones for which route enumeration, evaluation, skimming, loading and loading analyses are required can now be selected through the trip matrix to be assigned. Only zones with trips from or to them will be selected.

Intersection delays

Delays at intersections can be imported directly from Highway and included in transit line times.

Routing algorithm

Multiple access and egress connections between zones and lines are permitted as opposed to the single, least cost connection in earlier versions. Although some attempts are made to remove redundant connections, the onus is on the user to generate a sensible set of access and egress connections. Excessive connections to/from the public transport system could increase the numbers of routes enumerated/evaluated.

There is also a change in the transfer points available to the all-or-nothing process. Transfer points are generated in two stages: first with one or more GENERATE statements, and second by applying a set of rules to remove any extraneous ones.

In earlier versions, the all-or-nothing used transfer points generated by the first stage while enumeration used the final set. In 3.2, both use the final set. This should improve connectivity, although results from earlier versions will differ from those of this one.

Generally, with these changes, more routes will be enumerated than before with the same set of parameters. For those moving from 3.1, it would be worth re-examining the criteria for generating their non-transit network, and SPREAD and MAXFERS.


Version 3.1

Version 3.1Cube Voyager is the flagship Cube extension for forecasting urban, regional, and long distance passenger demand. Cube Voyager provides comprehensive forecasting functionality and replicates the functions of TP+ and TRIPS. To assist users of existing Citilabs products the following table shows approximately how programs correspond:

For customers with Cube ME, although the program can be used for gravity model distribution or growth factoring, its main purpose is maximum likelihood matrix estimation. Furthermore TRNBUILD is still available to assist those who are porting existing TP+ models into Cube Voyager.

Cube Voyager TP+ TRIPS

DISTRIBUTION TRIPDIST MVGRAM

FRATAR FRATAR MVGRAM

GENERATION TRIPGEN MVTEND MVMODL

HIGHWAY HWYLOAD MVHWAY (AVROAD AVCAP) AVJNET

MATRIX MATRIX MVMODL (MVMOD MVMNIP)

NETWORK HWYNET MVNET

PILOT TPMAIN

PUBLIC TRANSPORT TRNBUILD MVPUBM MVPUBG MVPUBL MVPUBR


Technical support and software updates

Technical support and software updatesThis section describes technical support and software update options. Topics include:

• Annual maintenance contract

• Technical support

• Software updates

Annual maintenance contract

Citilabs combines the services of technical support and software updates into one software maintenance service contract. Clients wishing to receive software maintenance services are encouraged to purchase an annual contract. These contracts may be purchased at any time. However, an update fee will be charged if a client wishes to resume the maintenance service that has been discontinued for a period of time. Please contact your local distributor or Citilabs for more information about the annual maintenance contract.

Technical support

Citilabs attempts to provide timely answers to questions about the behavior and performance of our software products covered under a software maintenance contract. You can contact Citilabs using:

• World Wide Web support

Submit your support query directly through our user support form at:

http://www.citilabs.com/cubeoutputsupportform.html

Submitting your support query through our Web-based support form allows your support request to enter our support request tracking and response system, where we assign it a tracking number and a support staff member for response.You


http://www.citilabs.com/cubeoutputsupportform.html


may also access the support form directly from the Help menu in Cube by selecting the ‘Submit Support Request to Citilabs’ item.

Citilabs maintains a World Wide Web site (www.citilabs.com) that hosts a variety of pages with up-to-date information about our products. Answers to general technical support questions will be posted on the Frequently Asked Questions (FAQ) pages as well. A direct link to the Citilabs home page is provided under the Help menu.

Citilabs maintains and monitors a user group at (http://citilabs.com/usergroup/). This site provides a forum for product users to post questions and get answers and information from other users as well as Citilabs staff. Notices for training and workshop events or other special Citilabs events are also posted to the user group. There is a Files area for the group to post examples and share information. Citilabs staff and clients have posted a number of useful example scripts to the files area of the group.

• E-mail support

Clients who licensed Citilabs software through a local distributor should e-mail support questions to the distributor. A current list of distributors can be found at http://www.citilabs.com/about-us/contact-citilabs

Clients who licensed Citilabs software directly from Citilabs should submit support queries using the web based support form described above. If you have problems accessing the Web submission form, then you may e-mail support questions including license number to [email protected].

Please do not send e-mail with large attachments (greater than 5 MB). Large attachments that need to be transmitted to Citilabs as part of a technical support issue should be uploaded to the Citilabs file transfer site (FTP). Detailed instructions for accessing the company FTP site are available at http://community.citilabs.com/citilabs/topics/ftp_site_information.



http://citilabs.com/usergroup/

http://citilabs.com/usergroup/

http://www.citilabs.com/about-us/contact-citilabs

http://www.citilabs.com/about-us/contact-citilabs

http://community.citilabs.com/citilabs/topics/ftp_site_information

http://community.citilabs.com/citilabs/topics/ftp_site_information


Software updates

You can update Cube Voyager in one of two ways:

• Replacement DVD

• Citilabs Web site

A replacement DVD, with the latest version of Cube Voyager, will be sent to a client upon the renewal of the annual maintenance contract. Clients should reinstall Cube Voyager from the replacement DVD to update the software as well as the maintenance contract expiration date.

Software updates can also be done from the Citilabs Web site (www.citilabs.com). Detailed instructions for Web-updating Cube Voyager are available at www.citilabs.com. A current maintenance contract is required for Web update.

WARNING: Do not attempt to update your software if your maintenance contract has expired, or you did not update the maintenance contract expiration date by re-installing from the replacement DVD. It will render your system inoperable.


http://www.citilabs.com/all_download.html

Cube Voyagercitilabs-website-resources.s3.amazonaws.com/resources/RS_CubeVoyager-6... · Cube...

Documents

Transcript of Cube Voyagercitilabs-website-resources.s3.amazonaws.com/resources/RS_CubeVoyager-6... · Cube...