Issues for numpy 2 and core package update #1617

• agasc Updates for numpy 2 agasc#206
• chandra_time Replace repr() with str() when passing to axTime3.convert_time chandra_time#61
• chandra_aca Changes for numpy 2 chandra_aca#197
• cxotime Stop setting copy to False by default cxotime#55
• cheta in fetch.get_interval, make sure db is closed cheta#286
• kadi
  • Add django 5.x compat for abstract class detection kadi#365
  • Change from deprecated np.chararray to np.char for encode kadi#370
  • Set Orbit event dict with Python native types not numpy types kadi#371
• mica Changes for Numpy 2 mica#324
• parse_cm
  • https://github.com/sot/parse_cm/pull/62
  • https://github.com/sot/parse_cm/pull/63 (based on 62)
• proseco Changes for numpy 2 proseco#411
• ska_numpy: Replace use of int_t with int64_t ska_numpy#17
• sparkles
  • Fix a test for a proseco fix and update a warning print for future numpy 2 compat sparkles#224
  • Replace np.in1d with np.isin sparkles#225
• xija
  • Check that DPA power validation in ska3-hope matches ska3-flight xija#151

Functional testing

• Confirm that kadi / django event database update written by ska3-hope is readable with ska3-flight
• Confirm xija_gui_fit works

Lessons learned

Numpy copy=False

Problems resulting in ValueError: Unable to avoid copy while creating an array as requested.

In general replace numpy copy=False with copy=None or remove it. Or else replace with np.asarray(...). See https://numpy.org/devdocs/numpy_2_0_migration_guide.html#adapting-to-changes-in-the-copy-keyword.

np.in1d is deprecated

Replace np.in1d with np.isin.

change in np.broadcast_arrays

The return type of np.broadcast_arrays used to be a list, and in numpy 2 it is a tuple. One fix is to just cast the output to a list.

change in numpy repr

repr() now includes the numpy type in the output string for numpy types.