Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 77.00%. Comparing base (2d42d74) to head (ebd354c).
Report is 5 commits behind head on main.

@@           Coverage Diff           @@
##             main     #507   +/-   ##
=======================================
  Coverage   76.99%   77.00%           
=======================================
  Files          66       66           
  Lines        8213     8216    +3     
=======================================
+ Hits         6324     6327    +3     
  Misses       1889     1889           

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

you need 2500 pixels because the Fov is 10"(for SCAO mode) and each pixel is equal to 4 mas. So why this figure is more than 4000 pixels ?

you need 2500 pixels because the Fov is 10"(for SCAO mode) and each pixel is equal to 4 mas. So why this figure is more than 4000 pixels ?

That 2500 is the offset (in pixels I think), the size is still 4096 (IIRC).

Also, I'm not sure how I created this particular figure. I shared it because it looks 'interesting' and seems to indicate a bug. The 'step' pattern is also seen in the original notebook, but it is way more pronounced with the updated code.

But note that this branch is still experimental, in particular because the tests don't pass.

Codecov says the new lines in detector_list.py are not covered. Do you think it would be worthwhile to add a test case that goes down that if statement? Especially given that this piece of code apparently produced a bug?

Ah thank you. Maybe the test fail exactly because the test code does not go through that path, but the notebooks work because they do take that branch. No that doesn't make sense. I don't really have time to look into it further now.

The notebook works! You need to change 10 to 0 for micado.cmds[ "DET.x"] (same for y) line 19

I suggest also to change cluster properties : mass=200 core_radius=0.005 and distance= 900 to see the same shape of cluster in each image.

The notebook works! You need to change 10 to 0 for micado.cmds[ "DET.x"] (same for y) line 19

Yes, setting DET.x to 0 should work, but that defies the purpose of that part of the notebook. It should be possible to place the DetectorWindow anywhere; in this particular notebook somewhere 'off axis' to show that the PSF changes over the field of view.

I'm not really inclined to spent too much time to continue working on this now though, because simulating the full detector array now seems to work 'fine'. Using the DetectorWindow is just for convenience so the simulation goes faster, but it is not essential.

So I hope you can continue your work @Camillechatenet by using the full detector array, or indeed as you did now the centered detector window.

Fixing #506 was essential though, because that prevented the FieldVaryingPSF to work at all. I'm not convinced the FieldVaryingPSF is working entirely as it should now, so maybe we should investigate that a bit more.

A good reason to fix this DetectorWindow problem anyway is that such bugs might be an indication of larger problems.

I fixed the failing test case. I also rebased this onto main, because this was a superset of another PR anyway. @hugobuddel, are you cool with me force-pushing that and then merging once it passed everything (which it does locally)? 🙂

I fixed the failing test case. I also rebased this onto main, because this was a superset of another PR anyway. @hugobuddel, are you cool with me force-pushing that and then merging once it passed everything (which it does locally)? 🙂

Apparently github is not smart enough to realize that the other PR has been merged. Maybe making a (dummy) commit would update the diff. Nevertheless feel free to force-push fix this, as I don't have enough time to work on this issue.