Roman Prism and Grism #416

robelgeda · 2021-02-24T18:03:39Z

This PR introduces the Prism and Grism modes for the Roman WFI based on cycle 8 design. It also introduces the ability to user override the aberrations files used to construct detectors.

This PR means #382 can be closed (any time).

Updates

WebbPSF

New Grism and Prism filters:
- GRISM_FILTER = 'G150'
- PRISM_FILTER = 'P120'
Changing filters to G150 or P120 changes the mode of the WFI and the aberrations files (unless there is a user aberrations override)
New WFI._aberrations_files: Dictionary with all default aberrations files paths. Dictionary key is the WFI mode.
New WFI._load_detector_aberrations(path): Class function to load aberrations files at path
New WFI. _get_filter_mode (wfi_filter): given a filter (string), wfi_filter, return the WFI mode. WFI modes are:
- Imaging
- Grism
- Prisim
New WFI.mode: Class property that returns the current mode of the WFI instance by passing the current filter to WFI. _get_filter_mode
New WFI.override_aberrations(aberrations_path): Overrides and locks the current aberrations with aberrations at aberrations_path. Lock means changing the filter/mode has no effect on the aberrations.
New WFI.reset_override_aberrations(): Releases WFI.override_aberrations lock and start using the default aberrations.
New Tests for mode and filter switching.
New Field point nearest point approximation (extrapolation).

WebbPSF Data

New wim_zernikes_cycle8_prism.csv and wim_zernikes_cycle8_grism.csv aberrations files.
Updated webbpsf-data/WFI/filters.tsv to add 'G150' and 'P120' (nlambda set to 20)
New Filter throughput files for G150 and P120 from Pandeia

Sample PSFs

codecov · 2021-02-24T19:13:39Z

Codecov Report

Merging #416 (476c6b9) into develop (5cdd41e) will increase coverage by 0.41%.
The diff coverage is 94.28%.

@@             Coverage Diff             @@
##           develop     #416      +/-   ##
===========================================
+ Coverage    64.06%   64.47%   +0.41%     
===========================================
  Files           11       11              
  Lines         4352     4408      +56     
===========================================
+ Hits          2788     2842      +54     
- Misses        1564     1566       +2

Impacted Files	Coverage Δ
webbpsf/roman.py	`90.72% <94.28%> (+0.72%)`	⬆️

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 5cdd41e...476c6b9. Read the comment docs.

…nto roman

mperrin

Looks good overall. See below for a few small requests for revisions but nothing big.

Note, I have not reviewed any of the data values within the new files, either for the aberration coefficients or the filter transmissions. I assume you have double checked those values are are they should be.

webbpsf/roman.py

mperrin · 2021-03-09T13:57:34Z

webbpsf/roman.py

+        # Update Aberrations
+        # ------------------
+        # Check if _aberrations_files has been initiated (not empty) and if aberrations are locked by user
+        if self._aberrations_files and not self._is_custom_aberrations:


Why the check for if the aberrations files are empty? Under what circumstances would that be the case?

When the class is initialized, with super(WFI, self).__init__("WFI", pixelscale=pixelscale), the filter is set to a new value in that function. We can not fill in self._aberrations_files before the super.__init__ call because the self._datapath attribute is not yet initialized. So we have to initialize self._aberrations_files to be empty and check for when super.__init__ sets the filter, so the code does not crash.

If we define self._datapath before super.__init__, this will not be needed. But I thought this was the safest way forward.

webbpsf/tests/test_roman.py

robelgeda · 2021-03-24T15:45:12Z

From @Skyhawk172 :

looking at the prism and grism spreadsheets, I noticed the “local_x” and “local_y” values are different than what we have in the “normal filter” spreadsheet.
For example, for our regular filters, these values range from -20.44 to +20.44, identically for all SCAs, whereas in the prism/grism spreadsheets they vary non-uniformally from SCA to SCA and go from about -25 to 17. The net effect of these differences is that the derived pixel value for some of these “local” coordinate values is negative and/or can be many tens of pixels from the edge of the detector, leading to an error “Could not get aberrations for input field point” when trying to get the aberration close to the edge of the detector.
This local-to-pixel conversion is done in _load_wfi_detector_aberrations.build_detector_from_table().
We make the assumption “to place (-20.44, -20.44) at (4, 4)“, but that means the prism/grism interpolator won’t cover all of the SCA pixels.

Skyhawk172 · 2021-06-15T17:20:52Z

webbpsf/roman.py

-from . import webbpsf_core
-from scipy.interpolate import griddata
+
+from scipy.interpolate import griddata, RegularGridInterpolator


RegularGridInterpolator is not needed anymore now that we've simplified the handling of the extrapolation case.

Skyhawk172 · 2021-06-15T17:21:17Z

webbpsf/roman.py

 from astropy.io import fits
 import astropy.units as u
 import logging

+from . import webbpsf_core
+from .optics import _fix_zgrid_NaNs


_fix_zgrid_NaNs is not needed anymore now that we've simplified the handling of the extrapolation case.

Skyhawk172

I've done a series of tests and all my results indicate that the implementation is correct. We will be examining the results closer to understand some interesting features (e.g. FWHM), but it does appear those features are the result of the reference data provided to us by GSFC.

robelgeda · 2021-06-22T14:09:21Z

Added data to v1.0.0rc2 and tests are passing

robelgeda added 3 commits February 24, 2021 12:59

add new modes (sepctral) for Roman WFI

fb65773

remove comment

9a8a80d

Doc strings and comments for WFI modes

38dac7f

robelgeda added 2 commits February 24, 2021 14:57

fix filter name (P120)

e03ce36

point travis to webbpsf-data-1.0.0.rc.tar

234df75

robelgeda added the Roman Affects Roman Space Telescope models in WebbPSF label Feb 27, 2021

robelgeda added 2 commits March 1, 2021 11:08

Merge branch 'develop' of https://github.com/spacetelescope/webbpsf i…

3f2c84b

…nto roman

add roman prism and grism tests

943a1a6

robelgeda marked this pull request as ready for review March 1, 2021 19:44

robelgeda requested review from Skyhawk172 and mperrin March 1, 2021 19:44

mperrin requested changes Mar 9, 2021

View reviewed changes

robelgeda added 4 commits March 9, 2021 12:11

spelling fix prisim -> prism

0cf595d

add roman doc-strings

02e45c7

fix roman doc-strings

90d82e5

make _load_detector_aberrations safe

f049e09

mperrin mentioned this pull request Mar 29, 2021

Proof of Concept: override_aberrations function for WFIRST (Roman) #382

Closed

field point nearest point approximation

e12cd40

robelgeda requested a review from mperrin May 3, 2021 17:37

robelgeda added 2 commits June 8, 2021 23:17

add RegularGridInterpolator to Roman

bda66c4

return to basic nearst extrapolation

ab699bf

Skyhawk172 reviewed Jun 15, 2021

View reviewed changes

Skyhawk172 approved these changes Jun 15, 2021

View reviewed changes

solve conflict

476c6b9

robelgeda force-pushed the roman branch from 88ae172 to 476c6b9 Compare June 22, 2021 03:10

ojustino mentioned this pull request Jun 24, 2021

Roman Cycle 9 Updates #466

Merged

14 tasks

robelgeda force-pushed the roman branch from cca1731 to 476c6b9 Compare July 13, 2021 17:28

mperrin approved these changes Jul 15, 2021

View reviewed changes

mperrin merged commit 9673b52 into spacetelescope:develop Jul 15, 2021

robelgeda mentioned this pull request Jul 23, 2021

Release 0.9.2 #471

Merged

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Roman Prism and Grism #416

Roman Prism and Grism #416

robelgeda commented Feb 24, 2021 •

edited

Loading

codecov bot commented Feb 24, 2021 •

edited

Loading

mperrin left a comment

mperrin Mar 9, 2021

robelgeda Mar 9, 2021 •

edited

Loading

robelgeda commented Mar 24, 2021

Skyhawk172 Jun 15, 2021

Skyhawk172 Jun 15, 2021

Skyhawk172 left a comment

robelgeda commented Jun 22, 2021

Roman Prism and Grism #416

Roman Prism and Grism #416

Conversation

robelgeda commented Feb 24, 2021 • edited Loading

Updates

WebbPSF

WebbPSF Data

Sample PSFs

codecov bot commented Feb 24, 2021 • edited Loading

Codecov Report

mperrin left a comment

Choose a reason for hiding this comment

mperrin Mar 9, 2021

Choose a reason for hiding this comment

robelgeda Mar 9, 2021 • edited Loading

Choose a reason for hiding this comment

robelgeda commented Mar 24, 2021

Skyhawk172 Jun 15, 2021

Choose a reason for hiding this comment

Skyhawk172 Jun 15, 2021

Choose a reason for hiding this comment

Skyhawk172 left a comment

Choose a reason for hiding this comment

robelgeda commented Jun 22, 2021

robelgeda commented Feb 24, 2021 •

edited

Loading

codecov bot commented Feb 24, 2021 •

edited

Loading

robelgeda Mar 9, 2021 •

edited

Loading