Fixed bugs in Star and prediction

[altairgomes]

• Now Star is able to fully receive astrometric parameters from the user.
• FIxed issue #18 Star now calculates the robust propagation of the position of the star and correspondent uncertainties.
• Fixed issue #27 Now Star is able to download and use the distance from Bailer-Jones et al (2018).
• Fixed issue #43 that in print(Star), pmDEC was printed wrong.
• prediction function now makes use of the user input of the star to calculate faster the occultation parameters.
• prediction now creates the time array inside each division to avoid memory overflow.
• prediction now propagates the position of the stars using only the proper motions before comparing the stars with the ephemeris.
• Partially Fixes issue #45 wherein occ_params, when the instant of the occultation was outside the given range. The function now gives appropriate error messages. The automatic range search was increased to 50 min from central instant in a recursive search.

[Bmorgado19]

Please, correct the docstring from Occultation(), now the time should be within 50 minutes range and not 10.

In Star() docstring please add if bjones default is True or False. Also, when bjones=True, please print a sentence similar to: Downloading star parameters from X/XXX/xxx, in the same manner as GDR2 and NOMAD.

[altairgomes]

I applied the suggested modifications, updated the changelog, and fixed a minor bug that happened when more than one star was found. This bug is because choice_star() require a "source" parameter stating which catalogue is being used for correct messaging. The parameter was not being passed by searchgaia() and getcolors().