search

aperiosoftware / astropy

Repository for the Astropy core package
www.astropy.org
BSD 3-Clause "New" or "Revised" License
0 stars 0 forks source link

More HCOMPRESS errors #59

Closed Cadair closed 1 year ago

Cadair commented 1 year ago

nothing_makes_sense

github-actions[bot] commented 1 year ago

👋 Thank you for your draft pull request! Do you know that you can use [ci skip] or [skip ci] in your commit messages to skip running continuous integration tests until you are ready?

Cadair commented 1 year ago

This PR should have failing tests, which appear to be as a result of quantize. I checked the decompress against fitsio and it returns the same numbers, so I think this is definately a compress side issue. Some more debugging also showed me this:

(Pdb++) scale
3.0142469963305665
(Pdb++) zero
6473046107.318646

(Pdb++) np.frombuffer(qbytes, dtype=np.int32)

array([-2147483594, -2147483637, -2147483559, -2147483590, -2147483613,

       -2147483603, -2147483589, -2147483614, -2147483637, -2147483625,

       -2147483556, -2147483608, -2147483559, -2147483560, -2147483599,

       -2147483595, -2147483609, -2147483579, -2147483635, -2147483567,

       -2147483628, -2147483598, -2147483571, -2147483599, -2147483631],
      dtype=int32)

so the quantize call is returning all default ZBLANK??

Cadair commented 1 year ago

This went green locally, so I think we should be all good.

If we can merge this I will then rebase the main PR and we should be good to go.