Fix ba mapping - Githubissues

This PR addresses multiple issues and questions raised by @chrklemm in the thread linked below:

Specific code changes include:

Revised the "tell.data_spatial_mapping" function to fix a bug we identified and to streamline the code.
Added docstrings to describe the two dataframes retuned by the "tell.train" function.
Added definitions and reference links for the statistical values used to evaluate the TELL MLP models. We note the suggestion to add a glossary of acronyms to the TELL package. This is a good suggestion and we will put it on our long-term development list.

Suggestions to improve the paper were addressed as follows:

Overall, the software seems to be very specific to the 54 Balancing Authorities mentioned above. Would it be possible to apply the algorithm to other areas, such as Europe? If so, instructions should be provided. We updated the title of the paper to reflect that TELL was designed to work in the United States. We also added a sentence to the paper and a section to our User Guide outlining how the model could be adapted to work elsewhere.
The first sentence is very long and could be separated after "CONUS". This sentence is now broken up.
The reference to the IM3 project is not mandatory. We opted to leave in the reference to IM3 as it provides context for how the model is intended to be used.
Could a statement be made about the extent to which "tell" could also be used for other regions than the 54 balancing authorities mentioned? For example in Europe? Added one line to the paper and a section to our User Guide explaining how the model could be applied elsewhere.
Line 73, "Output yearly 8760-hr time-series": How are leap years handled? TELL will predict a load for every hour in the meteorology time-series that it takes as input. In the case of weather forcing for the IM3 project, leap years produce an 8784-hr time-series of meteorology values so TELL will mirror that.
Fig. 1: The caption is on a different page than the figure itself. Since the formatting of the paper is controlled by the pdf builder this is out of our control.

IMMM-SFA / tell