These files contain the code (three executables in total) and the results carried out by that code. It provides the statistical power analysis and its sensitivity for Kolmogorov-Smirnov (KS), Anderson-Darling (AD) and the Modified Anderson-Darling (MAD) tests for Extreme Value (EV) distributions. The power is estimated via a Monte Carlo approach with 10^9 samples, under "case 0" conditions. The tests were performed varying the reference distribution, the sampling distribution, the sample size and the critical value. The sensitivity analysis was performed on the shape parameter of EV distribution.
Statistical power estimation in Extreme Value Theory
Federico Reghenzani;Giuseppe Massari;William Fornaciari
2018-01-01
Abstract
These files contain the code (three executables in total) and the results carried out by that code. It provides the statistical power analysis and its sensitivity for Kolmogorov-Smirnov (KS), Anderson-Darling (AD) and the Modified Anderson-Darling (MAD) tests for Extreme Value (EV) distributions. The power is estimated via a Monte Carlo approach with 10^9 samples, under "case 0" conditions. The tests were performed varying the reference distribution, the sampling distribution, the sample size and the critical value. The sensitivity analysis was performed on the shape parameter of EV distribution.File | Dimensione | Formato | |
---|---|---|---|
code.zip
accesso aperto
Descrizione: Source code
:
Publisher’s version
Dimensione
14.87 kB
Formato
Zip File
|
14.87 kB | Zip File | Visualizza/Apri |
power_test_results.zip
accesso aperto
Descrizione: Results of the power tests
:
Publisher’s version
Dimensione
27.04 kB
Formato
Zip File
|
27.04 kB | Zip File | Visualizza/Apri |
sensitivity_results.zip
accesso aperto
Descrizione: Results of the sensitivity tests
:
Publisher’s version
Dimensione
4.91 MB
Formato
Zip File
|
4.91 MB | Zip File | Visualizza/Apri |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.