performance_measures: Performance measures.
In jakobbossek/asr: Algorithm Selection in R

Performance measures which are frequently used in the field of algorithm selection in both the continuous and discrete domains.

getScore(times, score, ...)

getPARScore(times, cutoff = 3600, f = 10, ...)

getPQRScore(times, cutoff = 3600, f = 10, p = 0.5, ...)

getPERTScore(times, cutoff = 3600, f = 1, ...)

getSuccessRate(times, cutoff = 3600)

getFailureRate(times, cutoff = 3600)

getRuntimeQuantile(times, cutoff = 3600, p = 0.5, ...)

getHVScore(times, cutoff = 3600, p = 0.5, normalize = FALSE, ...)

`times`	[`numeric`] Vector of running times of algorithm runs.
`score`	[`character(1)`] Name of score.
`...`	[any] Arguments passed down to score functions.
`cutoff`	[`numeric(1)`] Cutoff time for algorithm runs in seconds. Runs with running times exceeding the cutoff time are categorized as failed. Default is 1 hour.
`f`	[`numeric(1)`] Penalty factor. Each failed run is penalized by `cutoff` times `f`. Default is 10, which is quite common in the literature on algorithm selection.
`p`	[`numeric(1)`] Quantile. Default is 0.5.