Skip to contents

Power for regression (or any continuous outcome)

Usage

rpower(rsq, n, k = 1, p = k, alpha = 0.05)

rsamsize(rsq, k = 1, p = k, power = 0.8, alpha = 0.05)

rpower_beta(rsq, beta)

Arguments

rsq

Fraction of variance explained by the features to be tested

n

Sample size

k

Number of parameters to test

p

Number of parameters in model (default: k)

alpha

Type 1 error rate

power

Desired power

beta

Beta coefficient

Examples

rpower(0.05, 40)
#> [1] 0.2812148
rpower(0.05, 40, 5)
#> [1] 0.1398391
rpower(0.05, 40, 5, 20)
#> [1] 0.1297087
rpower(0.05, 40, 5, 20)
#> [1] 0.1297087
rsamsize(0.05)
#> [1] 159

rpower_beta(0.1)
#> [1] 0.3333333
rpower_beta(beta=0.13)
#> [1] 0.01661914