Route Aggregation

[amandaha8]

• This is my first attempt at aggregating up the trip level statistics to route level based on Tiffany's work.
• I experimented with groupby to see if there was any difference with observed=True, observed=False, etc and try to make things faster.
• I wrote a function that checks the original trip level rows with the aggregated one row=one route results because some of the groupby combos yielded mistakes.
• I edited Tiffany's original code to exclude columns that are no longer needed/relevant.
• I added in new columns for the metrics derived from the trip level statistics such as how many minutes of the trip recorded 2+ GPS pings.
• I reran all the trip level statistics for the months because some columns are added later in the route aggregation step.
• This PR does not address trips that are outliers. For example, there are some trips in which the scheduled trip time is only 12 minutes but the difference between the first and last time stamp is 1000 minutes.

[github-actions]

nbviewer URLs for impacted notebooks:

• rt_scheduled_v_ran/05_fresh_nb.ipynb
• rt_scheduled_v_ran/06_vp_usable_exploration.ipynb

[tiffanychu90]

• Research Request - Trip Level to Route Level Metrics  #966