adds rcf ForecastLeadTimeBins, corrects misc. formatting

[samland1116]

• Adds ForecastLeadTimeBins() class to existing rcf's
  • allows for static or dynamic binning
  • offers examples of the various methods you can define dynamic bin dictionary in docstring
• Corrects various formatting issues (as encountered) and removes unused dependencies (as encountered)

Also could include:

• example in docs -- not sure how this translates over to Sam Lamont's docs update for v0.6?

[samlamont]

I'm getting the error: TypeError: ForecastLeadTimeBins._validate_bin_size_dict() takes 2 positional arguments but 3 were given when running rcf.ForecastLeadTimeBins() in the test_add_row_udfs test

[samland1116]

I'm getting the error: TypeError: ForecastLeadTimeBins._validate_bin_size_dict() takes 2 positional arguments but 3 were given when running rcf.ForecastLeadTimeBins() in the test_add_row_udfs test

resolved this by adding in the @staticmethod decorator -- merged in the newly updated main to consider other recently merged PR as well

[samlamont]

To me it might be a bit more clear to define the keys as just an integer (or maybe even string?) then the bin could be some time value (could also be a string?), for example:

bin_size = {
    "1": "6 hours",
    "2": "12 hours",
    "3": "1 day"
}

where bin 1 would be 0-6, 2 would be 6-18, etc. And the values (whether strings or datetime objects) could be converted to datetime objects is not already. Then the resulting field would just be the simple key value 1,2, or 3 rather than the time interval string, which could be hard to filter on if there's a long record of forecasts in the table

[samland1116]

Updated to consider inputs of format:

output:
     {
         "PT0H_PT6H",
         "PT6H_PT12H",
         "PT12H_PT1D",
         "PT1D_PT2D",
     }
     
     input:
     1. constant bin size, timedelta:
         pd.Timedelta(days=5)
         
     2. variable bin size, dict:
         [
             {start_inclusive: pd.Timedelta(hours=0), end_exclusive: pd.Timedelta(hours=6)},
             {start_inclusive: pd.Timedelta(hours=12), end_exclusive: pd.Timedelta(hours=18)}
         ]
         
     3. variable bin size, dict with string/datetime keys:
         {
             "bin1": {start_inclusive: pd.Timedelta(hours=0), end_exclusive: pd.Timedelta(hours=6)}
         }
         
         - allow string, datetime timedelta as well

[samlamont]

thanks for your efforts on this!