Prop.test {pairwiseCI}

Wrapper to prop.test(stats)

Description

Only for internal use in pairwiseTest.

Usage

Prop.test(x, y, alternative = "two.sided", test=c("prop.test", "fisher.test"), ...)

Arguments

x

a vector of success and failure in sample x, or a data.frame with a column of successes and a column of failures, then colSums are used.

y

a vector of success and failure in sample y, or a data.frame with a column of successes and a column of failures, then colSums are used.

alternative

character string defining the alternative hypothesis

test

a single character string: which function to be called, choices are "prop.test" for the chi-square test, and "fisher.test" for Fishers exact test, as they are defined in package stats

...

arguments to be passed to prop.test(stats) or fisher.test(stats)

Details

Just a wrapper function to call prop.test(stats). If x, y are data.frames containing two columns taken to be counts of successes and counts of failures, columnwise sums of x,y are calculated. The total number of successes and the total number of trials is then passed to prop.test.

Value

An object of class "htest", as defined by prop.test(stats)

See Also

prop.test, and pairwise.prop.test in stats

Examples

# If input is a data.frame:

set.seed(1234)

trials=rep(20,8)
success <- rbinom(n=8, size=trials,
 prob=c(0.2,0.2,0.2,0.2, 0.3,0.3,0.3,0.3))
failure <- trials-success

f<-as.factor(rep(c("group1", "group2"), each=4))

data<-data.frame(success=success, failure=failure, f=f)

g1<-subset(data, f=="group1")[,c("success","failure")]
g2<-subset(data, f=="group2")[,c("success","failure")]

g1
g2

# Prop.test calculates the columnwise sums and calls prop.test stats:

Prop.test(x=g1, y=g2)

# should be the same as:

CS1<-colSums(g1)
CS2<-colSums(g2)

CS1
CS2

prop.test(x=c(CS1[1], CS2[1]), n=c(sum(CS1), sum(CS2)))