Roman Cycle 9 Reference Data file corrupted

[Skyhawk172]

While looking more closely at the Zernike aberrations, I realized that the "wim_zernike_cycle9" data file used by WebbPSF does not match the original Cycle 9 file delivered by Goddard. See screenshot. It appears something went wrong when the original file got manually parsed for use by WebbPSF. @ojustino is currently investigating the root cause of how this happened.

I did a quick test by properly parsing the original file and the results are as expected (i.e. identical to what Goddard reports), so clearly the "wim_zernike_cycle9" file currently used by WebbPSF is wrong but the fix is straightforward.

[ojustino]

The root cause here was a mistake in copying values from Goddard's original spreadsheet for Roman's imaging mode to wim_zernike_cycle9_OG.csv. I remade it and ran it through the steps in Process_zern_and_filters.ipynb to generate a new wim_zernike_cycle9.csv file. @Skyhawk172 and I verified that the new file's Zernike values align with the spreadsheet's.

I also noticed that the aforementioned files, wsm_grism_zernikes_cycle9.csv, and wsm_prism_zernikes_cycle9.csv all had values from their corresponding spreadsheets' "RMS (nm)" columns where the CSV files' headers actually call for values from the "PV (waves)" columns instead. I remade all of these files using the proper columns.

These changes are now reflected in the ITAR storage location's webbpsf-data-source/WFI directory. I've also added a cycle9 directory to sources there so we have a mutually accessible record of the updates we made for Cycle 9 in #466.