SensSpec: Sensitivity and Specificity

In mathykathy26/kagRfun:

Description

Function to calculate a table of sensitivities, specficities and error rates for classification procedures using predicted probabilities and true classification labels.

Usage

SensSpec(x, grouping, thresh = seq(0, 1, 0.05), plots = T, table = F,
  legend.loc = c(0.7 * max(thresh), 0.85))

Arguments

x	Vector of predicted probabilities for classification. For example, the predicted probabilities from a logistic model.
grouping	The vector of true class labels for the data.
thresh	Vector of desired thresholds for classification. The table will have as many rows as thresholds.
plots	Logical, plot the sensitivity, specificity and error rate curves versus the threshold vector on one plot.
table	Logical, returns the table of sensitivities, specificities and error rates for all thresholds.
legend.loc	Vector of two values for the x and y coordinate of the legend, respectively. Applicable only when plot is TRUE.

Examples

test.dat<-data.frame(x=rnorm(10),y=rbinom(10,2,.5))
SensSpec(test.dat$x,test.dat$y)
SensSpec(test.dat$x,test.dat$y,plot=FALSE,tab=TRUE)

mathykathy26/kagRfun documentation built on May 29, 2019, 4:40 a.m.