Over-assignment in model

[goreaditya]

After fixing the household income in aa88b28 the model started over-assigning trips. Investigate this issue.

[jfdman]

Some specific summaries:

1. Total SD and LD trips by purpose for the latest run compared to pre-fix run. Did LD trips increase significantly?
2. Trip length frequency distributions and average distance by purpose for SD and LD model.
3. Investigate VOT in mode choice as first step to understanding trip length issue. The low VOT category has a much higher sensitivity to cost than mid and high income categories, which could lead to much less sensitivity to distance in destination choice (due to use of mode choice logsums as measure of accessibility).