binaryClassifierEffect: Binary Classifier Effects

In ivanliu1989/RQuant: A collection of generic functions

Description

Produce a list of ggplots, showing the relationship between each predictor variable and a probability.

Given a data frame and a vector of probabilities, such as those produced by a binary classifier model, this function produces a list of ggplots, showing the relationship between each of the named variables and the probability.

It produces a violin plot for factors, a scatter plot with a loess fitted curve for numeric, and an error otherwise.

Usage

binaryClassifierEffect(probs, df, var.names, max.n = 1000, display = FALSE,
  log.names = NULL)

Arguments

probs	A vector of estimated probabilities, from 0 to 1.
df	A data frame containing variables that were used to predict some binary outcome.
var.names	A vector of column names from df.
max.n	The maximum number of points to include in the resulting scatter plots.
display	A logical, that determines is the resulting plots are displayed or just returned.
log.names	A vector of column names to plot in log scale

Value

A list of ggplot objects, suitable for printing or saving.

See Also

qplot

Examples

df<-mtcars
df$am<-as.factor(df$am)
model<-glm(vs ~ am*mpg, data=df, family=binomial)
probs<-model$fit
binaryClassifierEffect(probs,df,c('am','mpg'))

ivanliu1989/RQuant documentation built on Sept. 13, 2019, 11:53 a.m.