plot.prm_power: Visualize the power analysis of a randomization test
In holub008/nymph: Randomization tests for nonparametric inference

Description Usage Arguments Author(s) Examples

If alpha is left unspecified, the distribution of p-values is visualized. If alpha is specified, power is plotted as a function of alpha

1 2	plot.prm_power(x, ..., statistic, alternative, alpha = seq(0.01, 0.25, by = 0.01))

`x`	a prm_power object to be plotted
`statistic`	the name of the statistic to be visualized
`alternative`	the alternative hypotheses for p-value determination. one of "less", "greater", "two.sided"
`alpha`	the numeric significance level(s) [0-1] used in inference - determines the FPR

kholub

gen_data <- function(){ data.frame(x = c(rnorm(50), rnorm(50, 1)), class = rep(c('a', 'b'), each = 50)) }
mcrp <- mcr_power(gen_data, c(mean = function(df){ mean(df[df$class == 'a', 'x']) - mean(df[df$class == 'b', 'x']) }, 
                              median = function(df){ median(df[df$class == 'a', 'x']) - median(df[df$class == 'b', 'x']) }),
    'class')
plot(mcrp, statistic = 'median', alternative = 'two.sided')
plot(mcrp, statistic = 'median', alternative = 'two.sided', alpha = NULL)