Generate ROC curve data

Share:

Description

Generate data suitable for ROC curve plotting from the results of run.tests

Usage

1
generate.roc(vals, pval = TRUE)

Arguments

vals

list, data structure as returned by run.tests

pval

boolean, whether the values in vals represent pvalues

Details

calculates the power via calculate.power for all significance levels from 0 to 1.

Value

array of values suitable for plotting via roc.plot

Author(s)

Sebastian D├╝mcke duemcke@mpipz.mpg.de

See Also

run.tests and roc.plot for plotting

Examples

1
2
3
4
5
noises <- cbind(lin=c(.1,.5,.8),circ=c(.2,.4,.6))
mycor <- function(...) cor(...)^2
results.cor <- run.tests(mycor,args=list(),types=c(1,7),noises=noises,nsim=100,size=50)
roc.data <- generate.roc(results.cor,pval=FALSE)
roc.plot(roc.data,legend=noises)