pypeit_setup error (fixed)

[igorandreoni]

A simple fix to an error I came across when running pypeit_setup. I am opening an issue instead of a PR because I am not sure if this would pass tests for all the older versions. I am using v1.16.0 of Pypeit on a Mac M1 laptop.

Command:
pypeit_setup -s soar_goodman_red -r /Users/igor/SOAR/data/raw/ -c all

Error:
Traceback (most recent call last): File "/Users/igor/miniforge3/envs/pypeit/bin/pypeit_setup", line 8, in <module> sys.exit(Setup.entry_point()) ^^^^^^^^^^^^^^^^^^^ File "/Users/igor/miniforge3/envs/pypeit/lib/python3.11/site-packages/pypeit/scripts/scriptbase.py", line 113, in entry_point cls.main(cls.parse_args()) File "/Users/igor/miniforge3/envs/pypeit/lib/python3.11/site-packages/pypeit/scripts/setup.py", line 147, in main pypeit_files = ps.fitstbl.write_pypeit(output_path=output_path, cfg_lines=ps.user_cfg, ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/Users/igor/miniforge3/envs/pypeit/lib/python3.11/site-packages/pypeit/metadata.py", line 1803, in write_pypeit pypeItFile.write(ofiles[j], version_override=version_override, File "/Users/igor/miniforge3/envs/pypeit/lib/python3.11/site-packages/pypeit/inputfiles.py", line 565, in write setup_lines = yaml.dump(utils.yamlify( ^^^^^^^^^^^^^^ File "/Users/igor/miniforge3/envs/pypeit/lib/python3.11/site-packages/pypeit/utils.py", line 1445, in yamlify elif isinstance(obj, (np.string_, str)): ^^^^^^^^^^ File "/Users/igor/miniforge3/envs/pypeit/lib/python3.11/site-packages/numpy/__init__.py", line 397, in __getattr__ raise AttributeError( AttributeError: np.string_was removed in the NumPy 2.0 release. Usenp.bytes_ instead.. Did you mean: 'strings'?

Fix:
I replaced elif isinstance(obj, (np.string_, str)): with elif isinstance(obj, (np.bytes_, str)): in utils.py

All works fine now.

[igorandreoni]

P.S. I had to make the same change in linetools/utils.py

[kbwestfall]

Thanks for the bug report @igorandreoni ! Can you confirm you're using numpy version 2 (or higher)? I expect this is because pypeit 1.16.0 will have minor bugs like this when used with numpy 2 (released after pypeit 1.16.0). If so, you should be able to avoid this by downgrading to, e.g., numpy==1.26.4. We hope to release pypeit version 1.17.0 within the next couple of weeks, which will address these numpy==2.x.x errors.

[igorandreoni]

Right, I am using numpy==2.0.1. Thanks for the advice @kbwestfall, I am looking forward to pypeit v1.17.0!