Transform biomarkers stored as wide to long

Multiple biomarkers measured on the same subjects are often stored as multiple columns in a data frame. This is a convenience function that transforms the data into long format, suitable for use with ggplot and geom_roc

Usage

melt_roc(data, d, m, names = NULL)

Arguments

data

Data frame containing disease status and biomarkers stored in columns

d

Column containing binary disease status. Can be a column name or index

m

Vector of column names or indices identifying biomarkers

names

Optional vector of names to assign to the biomarkers. If NULL, names will be taken from the column names

Value

A data frame in long format with three columns: D = binary disease status, M = biomarker value, and name = biomarker name

Examples

D.ex <- rbinom(50, 1, .5)
widedata <- data.frame(D = D.ex, M1 = rnorm(50, mean = D.ex, sd = 1), 
   M2 = rnorm(50, mean = D.ex, sd = .5))
longdata <- melt_roc(widedata, "D", c("M1", "M2"))
ggplot(longdata, aes(d = D, m = M, color = name)) + geom_roc()