fifer (version 1.1)

univariate.tests: Extract p values for a data frame

Description

Given a dataframe, this function predicts the specified categorical variable using each column in the dataset, one at a time. The function will automatically select whether to do a chi-square test, a t-test, or an ANOVA. See details.

Usage

univariate.tests(dataframe, exclude.cols = NULL, group, parametric = T, ...)

Arguments

dataframe

a data frame containing both the variables and the grouping variable

exclude.cols

a vector indicating (either numeric or character) which columns should not have a significance test

group

a string with the name of the grouping variable

parametric

Should parametric tests be used? Defaults to TRUE.

...

other arguments passed to t.test or wilcox.test

Value

a vector of p values

Details

Extract the p value from a univariate significance test univariate.tests will look at each column in the dataframe, then perform a t-test (or wilcoxen.test if parametric=TRUE), ANOVA (or kruskal.test if parametric=TRUE), or chi-square test where the grouping variable serves as the independent variable. The computer will chose a chi-square test of one of the following three conditions is met: (1) the variable is a factor, (2) the variable is a character variable, or (3) the variable has less than four unique values. An ANOVA (or Kruskall) will be used if the number of levels of the grouping variable is greater than two. In all other cases, a t-test (or wilcoxen) will be used.

Examples

k = data.frame(cbind(ID=1:100,
			A = rnorm(100),
			B = rnorm(100),
			C = rnorm(100),
			Group = rep(1:2, times=50)))
univariate.tests(dataframe = k, exclude.cols=1, group="Group")