Implement FCA and 2SFCA Functions

[JamesSaxon]

Follow the format laid out in access/access/fca/

• [x] Use the weighted_catchment method if possible.

• [x] Ultimately this should be tested (i.e., matching the earlier functions), as:

  from access import access
  
  a = access(...)
  a.fca1()
  a.fca2()
  a.access["fca2"] # this.

  This is the main priority for the week of June 24.

• [ ] Create Examples in documentation.

[20bryan]

Created a new branch (FCAand2SFCA) with the first round implementation of the base FCA and Two Stage FCA methods.

• changed documentation & method parameters/headers
• base FCA matches up exactly, 2SFCA has a slight discrepancy for certain values <-- to be addressed

[JamesSaxon]

Thanks Bryan, these look good. A couple of things --

• In fca1.py, I would prefer that the defaults for demand_index and supply_index remain True instead of geoid, since the latter is less general. As it is, the access class will try to do this -- that is, set the location names to the index.
• Thanks for creating the "normalize" option -- good idea. I see that for 2SFCA argument is there, but it's not implemented at the bottom, so please do remember that.
• The df.to_frame(name = ).join(df2.to_frame.... calls are a little verbose. I might suggest pd.concat -- something like

  s2 = pd.Series(index = range(10), data = range(0, 20, 2))
  s3 = pd.Series(index = range(10), data = range(0, 30, 3))
  df = pd.concat([s2, s3], axis = 1)

  although you could convince me that dealing with missing values in this case is enough messier to justify your way.