Marc Schefer

Sorry this was left unaddressed. We don't currently output the full covariance matrix, this is not that easy as the matrix is based on internal parameters. I'll classify this as...

Did we ever look into that?

Then we should for the version after this one.

Yes, I have fixed those "value outside range" to now be warnings and not stop the fit, the parameter value is clamped to the value you specified in the configuration....

I don't see the point of having a ticket for the warnings, they are working as intended. As for what reason the GSL fit is behaving weirdly in this case,...

It's weird, we test gsl as part of our automated tests so I think it's working in normal cases. Could it be possible those failure cases are related to what's...

To avoid the performance issues of calling the Python interpreter for every pixel we have implemented custom models as ONNX compute graphs, this doesn't require using machine learning techniques but...

@louisquilley Hi, I'm back sorry for the delay. Yes, the factor 0.59 is to be expected for an exponential since we set k to 1, I did not include it...

Good tests, thanks. I'm not really that surprised, I remember from early tests that results were very sensitive to initial values and how the parameters were set up. I didn't...

I think fast-math is a significant boost of performance but it comes at the price that we should not use NaNs and can't ever rely on them.