Open sfair opened 4 years ago
It seems this happens when you changes the environment of the use of astropy, for instance, from pip-based to conda-based. And it tries to read the cache whish is in another format.
Which version of Astropy are you using? They seem to already have made some modification concerning this warning: https://github.com/astropy/astropy/pull/9656.
Anyway. I think this should happen only once. Have you tried again to see if the warning reappears? Also, maybe try to clean the cache using astropy.utils.data.clear_download_cache() as it is suggested.
It's astropy v4.0.
It happens only once each time a notebook is run, and it's harmless.
clear_download_cache()
solves the problem.
After a clean install of SORA (v0.1) and Anaconda (v2020.02) on OSX Catalina, I got these two warnings:
WARNING: WrongDBMModuleWarning: Existing astropy cache is in an unsupported format, either install the appropriate package or use astropy.utils.data.clear_download_cache() to delete the whole cache; db type is dbm.gnu, but the module is not available [astropy.utils.data]
WARNING: leap-second auto-update failed due to the following exception: WrongDBMModule('Existing astropy cache is in an unsupported format, either install the appropriate package or use astropy.utils.data.clear_download_cache() to delete the whole cache; db type is dbm.gnu, but the module is not available') [astropy.time.core]
This happens to all provided exemples. The notebooks seem to run properly, though.
I'm unsure wether the code should address this issue or not.