Mismatch between Pyextremes and SciPy-generated return value plots

[Pranavesh-Panakkal]

I'm using Pyextremes to fit a Generalized Pareto Distribution (GPD) to my data. Then I used the learned parameters to generate a corresponding distribution object in SciPy. However, I noticed that the return value plot generated during the diagnostic plot (model.plot_diagnostic) does not match the plot produced by SciPy, and I am trying to understand what I might be missing.

# Extracting model parameters from Pyextremes
loc = model.distribution.fixed_parameters['floc']
c= model._fit_parameters['c']
scale = model._fit_parameters['scale']

Modeling GPD in Scipy.

genpareto(c=c, loc=loc, scale=scale)

Plot generated by Pyextremes

Plot generated by Scipy.

[georgebv]

I need more information before I can answer. How were the scipy graphs generated?