search

ejeschke / ginga

The Ginga astronomical FITS file viewer
BSD 3-Clause "New" or "Revised" License
120 stars 77 forks source link

numpy 2.0's change of copy semantics #1094

Closed ejeschke closed 3 months ago

ejeschke commented 3 months ago

Per the discussion in this astropy issue, ginga also has a number of uses of copy=False in numpy operations that will need to be changed to None to achieve the same semantics in numpy v2.x.

ejeschke commented 3 months ago

Fortunately, I think these are all internal, and not exposed in an API.

ejeschke commented 3 months ago

Good news. All of the uses of copy=False within ginga are with the numpy call astype(). According to the v2.0 doc the semantics in v2.0 are compatible with the semantics in v1.x.

So I don't think ginga core will need to change anything. External plugins in other packages could be affected, of course, depending on how they use numpy calls.

ejeschke commented 3 months ago

Closing. Will reopen if it proves to be an issue later.