calculateCoverage: Coverage probability evaluation

In forvis/forvision: Tools for forecast visualisation and evaluation

Description

Calculates estimates of actual coverage probabilities for the given nominal coverage probability.

Usage

calculateCoverage(af, pi, conf.level = 0.95)

Arguments

af	dataframe containing forecast data formatted using the Actual and Forecast Table Schema (AFTS), use showAFTS() to display schema specification details.
pi	nominal coverage probability (confidence level for the prediction interval to be assessed).
conf.level	confidence level for coverage probability estimates, default 0.95.

Details

Calculates coverage probability estimates for each method and each horizon contained in af. Coverage probability estimate is the percentage of cases when prediction interval contained actual. Confidence limits for the actual coverage probability are obtained using binom.test function (two.sided=TRUE).

To get a coverage probability graph, use plotCoverage().

Value

dataframe with coverage probability estimates for each method and each horizon contained in af.

Author(s)

Cuong Sai, Andrey Davydenko, and Maxim Shcherbakov.

See Also

plotCoverage

Examples

af <- createAFTS(example1_ts, example1_fc)
calculateCoverage(af, 90)

forvis/forvision documentation built on April 30, 2020, 3:28 a.m.