verification (version 1.41)

table.stats.boot: Percentile bootstrap for 2 by 2 table

Description

Performs a bootstrap on data from a 2 by 2 contingency table returning verification statistics. Potentially useful in creating error bars for performance diagrams.

Usage

table.stats.boot(CT, R = 100, alpha = 0.05, fudge = 0.01)

Arguments

CT

Two by two contingency table. Columns summarize observed values. Rows summarize forecasted values.

R

Number of resamples

alpha

Confidence intervals.

fudge

A numeric fudge factor to be added to each cell of the contingency table in order to avoid division by zero.

Value

• 2 row matrix with upper and lower intervals for bias, pod, far, ets.

See Also

table.stats

Examples

### example from Roebber. 	
RB1 <- matrix(c(95, 55, 42, 141), ncol = 2)
table.stats.boot(RB1, R = 1000   )