Analyzer: Add time-filtering functionality

[EwoutH]

This PR introduces a new time-filtering functionality for analyzing link and vehicle data over a specified time window (t_seconds). This allows for more granular, time-sensitive analysis, which is particularly useful when evaluating the most recent simulation behavior. It basically allows inputting t_seconds into vehicles_to_pandas, link_to_pandas and area_to_pandas, which makes it collect the last t_seconds of simulation time.

It's a clean implementation of PR #123 and is part of issue #120.

A few things to note:

• The t_seconds API is maybe still not ideal. Maybe you want to either:
  • Input t_start and t_end, and collect between that period.
  • Input a binwidth t_bin, and then slice up the data in equal parts of that width.
• It only touches the Analyzer class, so in theory can now be run at the end of a simulation instead of during it. For that the API above might make more sense.

Things to review:

• I think I'm triggering this at t=900 simulation time, but for some reason it takes t=930. Maybe the index is wrong by 1 or something like that
• I had to add the or t_seconds is not None to the if s.flag_pandas_convert == 0 or t_seconds is not None: check in vehicles_to_pandas. Might be more elegant ways to solve it.

This branch can be installed with:

pip install -U -e git+https://github.com/EwoutH/uxsim@pandas_time_filter#egg=uxsim

[EwoutH]

Okay, this is clearly something that can better be done with an time_bin parameter after analysis.

Back to the drawing board.

[toruseo]

If we choose time_bin approach, this PR can be closed? Just comfirming

[EwoutH]

Yes, succeeded by #130.