spacetelescope / PASTIS

Algorithm for analytical contrast predictions of coronagraphs on segmented telescopes
BSD 3-Clause "New" or "Revised" License
8 stars 4 forks source link

Pin hcipy to 0.4.0. #137

Closed ehpor closed 2 years ago

ehpor commented 2 years ago

Pins hcipy to 0.4.0, due to regression test failure found in #130. This should be unpinned when the reason for the regression has been found and fixed (or when a new PASTIS matrix is uploaded for the regression test).

An initial investigation showed that the regression was primarily numerical in nature, and than no physics changes were made. I could not tell you for sure what actually changed between 0.5.0 and 0.4.0. There were a lot of aperture changes so that would be my first choice.

This PR also moves both hcipy and progressbar2 to conda (conda-forge) installations rather than pip.