Matthew R. Becker
Matthew R. Becker
I think I hit this bug, but now even calling the save right after doesn't help. @tacaswell @SylvainCorlay @stonebig any ideas?
I only used `%matplotlib notebook` and saw this issue. I switched to `%matplotlib inline` and it went away.
Is this different compilers or different C++ stdlib ABIs?
So I have a PR open to handle the conda-forge issues via the pybind11-abi package: https://github.com/conda-forge/pybind11-feedstock/pull/79 The first item is out of my wheel house.
https://gist.github.com/beckermr/870414e5f168d3aec7f0d9e1c91adccf
I'm happy to code this up if there is interest. cc @rmjarvis @rearmstr @gbernstein @esheldon
I changed the issue title since it confused me. :) FWIW, as far as I know, this package aims to be compliant with the FITS standard and uses cfitsio internally,...
> According to https://heasarc.gsfc.nasa.gov/docs/software/fitsio/c/c_user/node76.html#ffpscl, it those cfitsio routines do not need to get set once and then set back again; it only affects the data read in. I read the...
My read is that we would either have to force the whole file to be read this way, or we need to set and then unset these parameters when reading...
We might be better off parsing the header values ourselves and adjusting the output by hand.