| critVal.power {geneRxCluster} | R Documentation |
critical region cutpoints
critVal.power(k, p0, target, pwr = 0.8, odds = 7)
k |
- window width(s) |
p0 |
- length 2 probabilities |
target |
- false discoveries wanted |
pwr |
- desired power |
odds |
- alternative odds ratio |
This version uses power and TFD and will limit windows screened
list of cutoffs and attributes
Charles Berry
gRxCluster for how and why this function is
used
# symmetric odds:
crit <-
critVal.power(5:25,c(1,1),5,pwr=0.8,odds=7)
crit[[1]]
sapply(crit,c)
# 5:1 odds
asymmetric.crit <-
critVal.power(5:25,c(1,5),5,pwr=0.8,odds=7)
# show the critical regions
par(mfrow=c(1,2))
gRxPlot(crit,method="critical")
gRxPlot(asymmetric.crit,method="critical")
rm(crit,asymmetric.crit)