Mention Table interface in "Creating a New Table File" in FITS docs

[astrofrog]

When people want to create a simple binary FITS table with no other HDUs, most users should probably just use the Table interface in Astropy instead and then write to FITS. Therefore I think in the io.fits docs (e.g. here: http://docs.astropy.org/en/stable/io/fits/index.html#creating-a-new-table-file) we should add a note to mention the Table interface, or maybe even provide an example.

(This is not high priority, so could be done as a starter GSoC project)

[embray]

👍 (This is practically a FAQ).

[vinayak-mehta]

@astrofrog Can you review this? https://github.com/vortex-ape/astropy/commit/9ba1f2db49e2455557f883903685c9a73125f15d

[astrofrog]

@vortex-ape - what you have is a good start, but please open a pull request so that we can comment on it in detail (pull requests can be updated once opened)

[embray]

I don't think this should have been labeled package-novice now that I look.

[embray]

Per #3561, it would also be nice to clarify in the docs that even when using the "lower-level" interface it is not strictly necessary to go through the effort of creating Column objects manually--that data already in a Numpy array can be used to directly create a FITS table as long as it's in a structured array. Direct use of the old interface is more of a FITS-ism, and only required in certain cases.

[vinayak-mehta]

@embray - What more benefits can I add?