plotImportance: Plot feature importance

In jasenfinch/metabolyseR: Methods for Pre-Treatment, Data Mining and Correlation Analyses of Metabolomics Data

Plot feature importance

Description

Plot Univariate or random forest feature importance.

Usage

plotImportance(x, ...)

## S4 method for signature 'Univariate'
plotImportance(x, response = "class", rank = TRUE, threshold = 0.05)

## S4 method for signature 'RandomForest'
plotImportance(x, metric = "false_positive_rate", rank = TRUE)

## S4 method for signature 'list'
plotImportance(x, metric = "false_positive_rate")

Arguments

x	S4 object of class Univariate or RandomForest
...	arguments to pass to specific method
response	response results to plot
rank	rank feature order for plotting
threshold	explanatory threshold line for the output plot
metric	importance metric to plot

Examples

library(metaboData)

x <- analysisData(abr1$neg[,200:300],abr1$fact) %>%
       keepClasses(cls = 'day',classes = c('H','1','5')) %>% 
       occupancyMaximum(cls = 'day') %>%
       transformTICnorm()
       
rf <- randomForest(x,cls = 'day')

plotImportance(rf,rank = FALSE)

jasenfinch/metabolyseR documentation built on Sept. 18, 2023, 1:25 a.m.