TransCAD generated OMX skims are not read properly in DaySim

[nsdhakar]

The OMX format skim files generated by TransCAD are read incorrectly by DaySim. Used the GISDK command "CopyMatrix()" to convert TransCAD format skims into OMX format in TransCAD 9.0. DaySim reads the OMX files successfully, however, the values are read incorrectly (very large numbers) and also do not correspond to correct OD cells in skims. Note that when opened in TransCAD or an OMX viewer or a programming language, the OMX skim files look fine. But for some reason, DaySim is not able to read them correctly.

A workaround is to read TransCAD generated OMX files in a programming language (e.g., R) and re-write back in OMX format. The regenerated OMX skims are read fine in DaySim.

DaySim has been reading OMX files generated from Cube, Emme, or a programming language (e.g., R, python) without issues. This points to a potentially issue in the way TransCAD is generating OMX files. Something to coordinate with Caliper.

[jfdman]

Have you reached out to Caliper about this?

…

On Fri, Aug 18, 2023, 2:07 AM Nagendra S Dhakar ***@***.***> wrote: The OMX format skim files generated by TransCAD are read incorrectly by DaySim. Used the GISDK command "CopyMatrix()" to convert TransCAD format skims into OMX format in TransCAD 9.0. DaySim reads the OMX files successfully, however, the values are read incorrectly (very large numbers) and also do not correspond to correct OD cells in skims. Note that when opened in TransCAD or an OMX viewer or a programming language, the OMX skim files look fine. But for some reason, DaySim is not able to read them correctly. A workaround is to read TransCAD generated OMX files in a programming language (e.g., R) and re-write back in OMX format. The regenerated OMX skims are read fine in DaySim. DaySim has been reading OMX files generated from Cube, Emme, or a programming language (e.g., R, python) without issues. This points to a potentially issue in the way TransCAD is generating OMX files. Something to coordinate with Caliper. — Reply to this email directly, view it on GitHub <#317>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ABHJIC7ORRS2TK3E37TWGMLXV4WFHANCNFSM6AAAAAA3VHOIFU> . You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>

[nsdhakar]

@jfdman - yes, sent an email to Vince.