Setting Starting Values for Fitting

[bliblabl]

I have a lot of appreciation for you work and really enjoy using it - thanks!

I had to make a scipy.stats.rv_continuous method myself and to fit it I need to set the starting values manually (arg1, arg2, arg3 from here: https://docs.scipy.org/doc/scipy/reference/generated/scipy.stats.rv_continuous.fit.html#scipy.stats.rv_continuous.fit). Is it already possible to do this via fit_model? If not, would it be possible to integrate this in pyextremes?

Would be really helpfull!

[georgebv]

It is not currently possible and I am not sure what would be the use case for this. This would only be applicable to the MLE model because Emcee doesn't even use scipy's fit method.

Regarding your problem, I think you can hardcode those in the _fitstart(data) method as specified in the scipy docs:

def _fitstart(self, data):
    return arg1, arg2, arg3

where argx is a hardcoded value

[bliblabl]

Thanks a lot. I did that and it worked (before I overloaded the _fit() method and that would have worked as well for the MLE).