added pandas day alias to fix problem when using hercules with resolutions < 1 month

[Jeffrey-P-McAteer]

Not sure where to fix this but I was getting the following errors when running with values of $DAYS < 30

./hercules --sampling=$DAYS --granularity=$DAYS --burndown --first-parent --pb "$OLD_CWD" | labours -f pb -m burndown-project --output stack_plot.png

Running: burndown-project
         Ratio of survived lines
2 days                  0.947440
4 days                  0.917328
6 days                  0.915297
8 days                  0.915124
10 days                 0.915124
12 days                 0.915124
14 days                 0.915124
14 days                 0.915124
resampling to year, please wait...
100%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 7/7 [00:00<00:00, 5108.77it/s]
too loose resampling - by year, trying by month
resampling to month, please wait...
Traceback (most recent call last):
  File "/usr/bin/labours", line 10, in <module>
    sys.exit(main())
  File "/usr/lib/python3.7/site-packages/labours/cli.py", line 449, in main
    modes[mode]()
  File "/usr/lib/python3.7/site-packages/labours/cli.py", line 204, in project_burndown
    interpolation_progress=True,
  File "/usr/lib/python3.7/site-packages/labours/modes/burndown.py", line 344, in load_burndown
    header, name, matrix, "month", report_survival=False
  File "/usr/lib/python3.7/site-packages/labours/modes/burndown.py", line 347, in load_burndown
    raise ValueError("Too loose resampling: %s. Try finer." % resample)
ValueError: Too loose resampling: M. Try finer.

the pandas.date_range freq argument accepts a number of different time strings including D for day resolution, which is what I wanted for a small project of mine.

[vmarkovtsev]

Thanks for a great contribution!

[vmarkovtsev]

Hey @Jeffrey-P-McAteer could you please fix the DCO and the CI so that I can merge this? I believe that a git rebase - -signoff should solve both problems.