I've just run `imgCloseLoop` yesterday on NCSA with `ts_phosim` on master (1.3.1), `ts_wep` on master (1.6.2) , `phosim_utils` on master (0.2.4), and it converged. Te-Wei Tsai , do you mean to test the `skySim` functionality?
When I run `imgCloseLoop` I provide the input sky file:
python bin.src/imgCloseLoop.py --inst comcam --numOfProc 25 --boresightDeg 0.03 -0.02 --skyFile tests/testData/sky/skyComCam.txt --output /project/scichris/aos/perturbations/imgCloseLoop_0-05/
So here skySim is used only to read the ra,dec from file https://github.com/lsst-ts/ts_phosim/blob/e25ac907d0385b109caf1ee4a803d9f248b54096/python/lsst/ts/phosim/SkySim.py#L126-L154
If I run it in that mode, I don't think it uses the SkySim functionality of converting ccd position to ra,dec , which is considered in the failing test (https://github.com/lsst-ts/ts_phosim/blob/e25ac907d0385b109caf1ee4a803d9f248b54096/tests/test_skySim.py#L125-L142 ) . So to test the conversion of ccd x,y to ra,dec , I could do something like in this notebook https://github.com/suberlak/AOS/blob/main/AOS_letter_simulate.ipynb . Te-Wei Tsai , is this what you had in mind ?
By the way, looking deeper into the issue, I would have thought that this test https://github.com/lsst-ts/ts_phosim/blob/e25ac907d0385b109caf1ee4a803d9f248b54096/tests/test_opdMetrology.py#L214-L226 should also fail since it tests almost identical functionality (ccd x,y --> ra,dec). But upon inspection the difference is that
So I think that basically given the `WcsSol.py` update , and that `ts_phosim` skySim uses it to get x,y --> ra,dec , it is good to update the test to reflect that. Another question : why are we using two different functions for such similar functionality?