spectra position identification

[lpagani91]

I met two problems with the finalized list of gaussians returned by gausspy+:

1. the CROTA value in the fits header of my data was ignored and therefore the absolute coordinates of the pixels got wrong.
   Of course, one can provide data only in the Ra-Dec frame or Galactic frame with no rotation but sometimes it is not convenient.

2. the pixel numbering differs by 1 unit between the file and the fits file (most - all ? - fits handlers outside Python, like ds9, count pixels starting from 1 and not from 0). I suggest that an option be set in the gausspy+.ini file to decide whether the user wants a pixel count starting from 0 or from 1.