First Stable Extreme Correction Class

[viccollado]

Add Extreme Correction Technique for sampled dataset based on the SC developed by R. Mínguez, V. Collado and F. J. Méndez.

Notes:

• At the moment the ExtremeCorrection class is only applied in the complete dataset (without dividing by WT). It has parameters that do not affect the model (for example, bmus) in the future it will be added to be able to apply the correction directly by Weather Types.
• If any want to apply the correction by WT now, the easiest way is split the datasets by WT and setting the variable as 0 if the value does not correspond to this WT. Contact: @viccollado for more info.

Important functions/classes added:

• ExtremeCorrection (extreme_correction.py): Apply the extreme correction class (see notebook example in Bluemath).
• OptimalThreshold (pot.py): Obtain optimal threshold for GPD based on studentized residuals (see R. Mínguez 2025).
• GPDPoiss model (pareto_poisson.py): Functions of GPD-Poisson model for AM of POT.
• Utils functions: add utils files used in some functions and classes.

[tausiaj]

Checked!