eva (version 0.2.6)

gevrEd: GEVr Entropy Difference Test

Description

Goodness-of-fit test for GEVr using the difference in likelihood between GEVr and GEV(r-1). This can be used sequentially to test for the choice of r.

Usage

gevrEd(data, theta = NULL)

Arguments

data

Data should be contain n rows, each a GEVr observation.

theta

Estimate for theta in the vector form (loc, scale, shape). If NULL, uses the MLE from the full data.

Value

statistic

Test statistic.

p.value

P-value for the test.

theta

Estimate of theta using the top r order statistics.

Details

GEVr data (in matrix x) should be of the form \(x[i,1] > x[i, 2] > \cdots > x[i, r]\) for each observation \(i = 1, \ldots, n\). The test uses an asymptotic normality result based on the expected entropy between the GEVr and GEV(r-1) likelihoods. See reference for detailed information. This test can be used to sequentially test for the choice of r, implemented in the function `gevrSeqTests'.

References

Bader B., Yan J., & Zhang X. (2015). Automated Selection of r for the r Largest Order Statistics Approach with Adjustment for Sequential Testing. Department of Statistics, University of Connecticut.

Examples

# NOT RUN {
# This will test if the GEV2 distribution fits the data.
x <- rgevr(100, 2, loc = 0.5, scale = 1, shape = 0.5)
result <- gevrEd(x)
# }