search

pcdshub / camviewer

LCLS Camviewer Application
0 stars 4 forks source link

Add lineouts and (super)gaussian curve fits. #8

Closed mcb64 closed 1 year ago

klauer commented 1 year ago

Is this ready for review?

mcb64 commented 1 year ago

Right, didn't assign anyone, did I?

klauer commented 1 year ago

So glad to see flake8 re-added 👍

This flake8 config might be of interest - we use it in other projects to make things a bit more lenient: https://github.com/pcdshub/cookiecutter-pcds-python/blob/master/%7B%7B%20cookiecutter.folder_name%20%7D%7D/.flake8

mcb64 commented 1 year ago

Yeah, I had figured out the E501, which is why I took out flake to begin with. Might as well make things standard though...

ZLLentz commented 1 year ago

Somewhat unrelated to this code itself: is there any easy way to feed generated images to camviewer to check out the curve fitting? We're a bit lacking in real beam to try gaussian fits on.

mcb64 commented 1 year ago

Not any more. I used to be able to use random XTC files, but I removed this.

When I tested, I typically just used a bright spot on a random cam image and zoomed in the ROI. I trust lmfit is doing the right fit... the only question is if I am displaying it properly, and it certainly looks like something is being fitted.

mcb64 commented 1 year ago

Just FYI, the jira ticket (https://jira.slac.stanford.edu/browse/LCLSECSD-1752) has a lot of fitted screenshots using this approach.

ZLLentz commented 1 year ago

I agree, in those images it 100% looks like it's doing the right thing