Hips2fits new service

[bmatthieu3]

Hi @bsipocz @adamginsburg @tboch

I have implemented a new package responsible for querying hips2fits: http://alasky.u-strasbg.fr/hips-image-services/hips2fits
This service take a HiPS name + astropy WCS and return a FITS/JPG/PNG image cutout of the HiPS for that WCS. The list of valid hips name is here: http://aladin.unistra.fr/hips/list
There are two methods:

• query_with_wcs that accepts a HiPS name + an astropy.wcs
• query doing the same but accepts different parameters instead of an astropy wcs (the center of projection, fov angle, projection, rotation...)

I added a simple doc containing 2 examples with the corresponding output plot images that can be obtained thanks to hips2fits. Some tests (remote too) have also been added.

Thank you very much for your reviews.

[pep8speaks]

Hello @bmatthieu3! Thanks for updating this PR. We checked the lines you've touched for PEP 8 issues, and found:

There are currently no PEP 8 issues detected in this Pull Request. Cheers! 🍻

Comment last updated at 2020-11-24 14:55:14 UTC

[keflavich]

This looks really good. Could you fix up the PEP formatting, though? I found a few small corrections that need to be handled.

[keflavich]

you may not be able to rely on this being present; if a wcs was created not from a fits Header (i.e., from scratch), it won't have this attribute populated

[bmatthieu3]

The hips2fits service asks for the size of the result image (width and height), it is mandatory. I changed the code, if the wcs has no pixel shape (pixel_shape None) then I raise an AttributeError saying either to give your wcs the pixel size of the image or to use the other query method

[bmatthieu3]

50f14de

[bmatthieu3]

Hi @adamginsburg,

I have seen a test is failing because it runs tests from the doc api (https://travis-ci.org/github/astropy/astroquery/jobs/695927179) which are remote tests. How to say these doc tests are remote and thus can be skipped ? Thanks

PS: I also added a changelog

[keflavich]

Add __doctest_skip__ = ['Hips2FitsClass.*'] at the top of the code that has Examples in docstrings (see, e.g., Vizier for examples)

[keflavich]

I don't have permission to push to this branch. It needs a simple rebase to handle changelog conflicts.

[bsipocz]

@keflavich - Does this need another round of review, or mergeable pending resolving the conflict and getting an all green CI?

[keflavich]

afaict, this was ready to go once those minor fixes (changelog, rebase, CI pass) were completed.

[bmatthieu3]

Hi @keflavich @bsipocz,

I added __doctest_skip__ = ['Hips2FitsClass.*'] to the beginning of the core.py file so that doc tests do not execute (because they are remote).
I do not see the CI tests running.
We are very excited to merge this into the next version of astroquery. Let me know what I should do to make this merged.

[keflavich]

I'm going to merge this on the command line - we need to put the changelog entry into the right section, but otherwise it's good to go.

[keflavich]

The remote tests pop up a blocking imshow - we should avoid that! However, the tests all passed.

[bsipocz]

@tinumide - when we get here, we should also look for doctest skips like this in the code files, and turn on the remote testing for them.

[tinuademargaret]

Alright, noted

[bsipocz]

@bmatthieu3 @keflavich - Please open a follow-up PR for fixing the imshow issue or skip the test, as we want to turn back on the remote CI testing soonish.

[bsipocz]

Retrospect I think this should have been a new module under cds, as part of our upcoming restructure.