tsftest.pow: 'tsftest' power function

In olli0601/abc.star: abc.star

Description

Compute the power of the two-sided one-sample multivariate equivalence test for population means of multivariate normal summary values with unknown population variance.

Usage

tsftest.pow(rho, cl, cu, n.of.y, p, support = c(0, Inf), log = FALSE,
  norm = 1)

Arguments

rho	Vector of quantiles
cl	Lower boundary point of the critical region
cu	Upper boundary point of the critical region
n.of.y	Number of replicate simulations
p	Number of variables
support	Support of the truncated power function (vector of dimension 2).
log	If TRUE, the power function is returned on the log scale.

Note

The power function can be truncated to support.

References

http://arxiv.org/abs/1305.4283

Examples

# power function of the two-sided F-test, to test equality of means for multivariate
# normal samples with unknown covariance matrix

n		<- 10
p		<- 3
cu		<- 8
rho 	<- seq(0, 2, length = 1024)
tmp <- lapply(c(0.01, 0.5, 1, 2, 3, 4, 5, 6, 7), function(cl)
		{
			data.table(cl = as.factor(cl), cu=cu, rho = rho, power = tsftest.pow(rho, cl, cu, n, p))
		})
tmp	<- do.call('rbind', tmp)
pp <- ggplot(tmp, aes(x = rho, y = power, colour = cl, group = cl)) + geom_line() + labs(y = 'Power\n(ABC acceptance probability)')
print(pp)

olli0601/abc.star documentation built on May 24, 2019, 12:53 p.m.