Diagnostic plots

This page demonstrates how to use the diagnostic plots from the ExtremePlots.jl package with the data structures provided by Extremes.jl.

Setup

Load the required packages:

using Distributions, Extremes, ExtremePlots, Gadfly

Simulate data from a Generalized Extreme Value (GEV) distribution:

pd = GeneralizedExtremeValue(0,1,0)
y = rand(pd, 300);

Fit the block maxima model by maximum likelihood:

fm = gevfit(y)

MaximumLikelihoodAbstractExtremeValueModel
model :
	BlockMaxima{GeneralizedExtremeValue}
	data :		Vector{Float64}[300]
	location :	μ ~ 1
	logscale :	ϕ ~ 1
	shape :		ξ ~ 1

θ̂  :	[-0.01753299188548797, 0.026174238797756186, 0.024881554077077973]

Histogram Plot

Displays a histogram of the data with the distribution overlay.

ExtremePlots.histplot(fm)

Probability-Probability (PP) plot

Compares empirical and model probabilities.

ExtremePlots.probplot(fm)

Quantile-Quantile (QQ) plot

Compares empirical and model quantiles.

ExtremePlots.qqplot(fm)