quantal size count_weight_gamma parameter is not working as expected

[raacampbell]

Setting this to 1 in Python allows the BScope2__2025_02_28 dataset to use the initial part of the curve and get a nice quantal size if around 10. In MATLAB this just is not working the same way. The curve looks wrong. So negative photon counts.

EDIT: the issue is partially resolved (I set the count_weight_gamma to 2.0). But I don't follow why 1.0 is not correct, as it is in Python.

[raacampbell]

The weights vector looks as expected when the user is asking for flat vs not flat. So the question is what the fit command itself is doing.

[raacampbell]

In Python, setting the count_weight_gamma value to 1.0 is needed to get a good fit with these data (see top image). The lower is 0.2 and the fit is bad.

[raacampbell]

With MATLAB this also works, but we need to set count_weight_gamma to 2.0 instead:

count_weight_gamma=2.0

count_weight_gamma=1.0

count_weight_gamma=0.2