produce.table1: Produce a "Table 1" for many clinical research reports.

In jraffa/MIMICbook: R functions for the Secondary Analysis of Electronic Health Records Book/ Course

Description

This function computes useful summary statistics from a data frame of numbers and factors into what many call a "Table 1" for many clinical research reports used to describe the study population. It can also be used to construct comparison tables for treatment groups or other strata.

Usage

produce.table1(x, labels = NULL)

Arguments

x	a data frame of patient characteristics from a study. All components in x are assumed to be numeric or factors.
labels	a list of reference values to use for factor variables. If NULL reports the most prevalent group as the referenced group. It will specify which group in the table.

Details

This function computes mean and standard deviations for numeric columns and number of individuals with a characteristic along with percentage for factor columns contained in a data.frame. Outputs a matrix of characters which can then be passed to some other functions for display in a research report. See the tableone package for a more comprehensive function that has many other bells and whistles.

Value

a character matrix of 1 column with means and standard deviations calculated for numeric columns and N and

Author(s)

Jesse D. Raffa

See Also

kable xtable tableone

Examples

require(knitr)
set.seed(1)
N <- 100;
dataf <- data.frame(age=runif(N,0,100),sex=sample(c("M","F"),N,replace=TRUE),cholesterol=rnorm(N,100,30))
out <- produce.table1(dataf,labels=c(NA,"F",NA))
kable(out)
dfbysex <- split(dataf,dataf$sex)
out1 <- produce.table1(dfbysex[[1]])
out2 <- produce.table1(dfbysex[[2]],labels=attr(out1,"labels"))
outbysex <- cbind(out1,out2)
colnames(outbysex) <- paste(colnames(outbysex), c(as.character(dfbysex[[1]]$sex[1]),as.character(dfbysex[[2]]$sex[1])))
kable(outbysex)

jraffa/MIMICbook documentation built on May 19, 2019, 11:51 p.m.