Use array-API-compatible dtype conversion in Combiner._weighted_sum

Fixes #924. The string-based weights.astype('float64') triggers a
DeprecationWarning in JAX's array API implementation, which is promoted
to a test failure in the py313-jax CI job.

Use xp.astype(weights, xp.float64) instead, where xp is the array
namespace already in scope at that point in the function. This works
across all array-API backends (numpy, jax, cupy, etc.) without
deprecation warnings.

The issue body also notes a 'stale' test in test_image_collection.py
(test_generator_ccds_without_unit) where it claims ImageFileCollection.ccds()
no longer raises ValueError. Verified locally against astropy 7.2.0 that
the test still passes as written — ccds() does still raise ValueError
when no unit is supplied (via CCDData.__init__ enforcing
_config_ccd_requires_unit). The 'stale test' part of the issue appears
to be based on a misreading of the current behavior, so the test is
left as-is.

No new tests needed: the JAX CI job is the verification for the array-API
dtype change.

Author: karlhillx

CHANGES.rst

@@ -1,3 +1,14 @@
+2.5.2 (unreleased)
+------------------
+
+Bug Fixes
+^^^^^^^^^
+
+- Fix dtype conversion in ``Combiner._weighted_sum`` to use the array-API
+  namespace form ``xp.astype(weights, xp.float64)`` instead of the deprecated
+  string-based ``.astype("float64")``. This resolves the ``DeprecationWarning``
+  that was being promoted to a test failure in the JAX CI job. [#924]
+
 2.5.1 (2025-07-05)
 ------------------
 

ccdproc/combiner.py

@@ -600,7 +600,7 @@ def _weighted_sum(self, data, sum_func, xp=None):
         # Turns out bn.nansum has an implementation that is not
         # precise enough for float32 sums. Doing this should
         # ensure the sums are carried out as float64
-        weights = weights.astype("float64")
+        weights = xp.astype(weights, xp.float64)
         weighted_sum = sum_func(data * weights, axis=0)
         return weighted_sum, weights