plotSupervisedRF: Supervised random forest MDS plot

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

Supervised random forest MDS plot

Description

A multidimensional scaling (MDS) plot of supervised random forest analysis

Usage

plotSupervisedRF(
  x,
  cls = "class",
  rf = list(),
  label = NULL,
  shape = FALSE,
  ellipses = TRUE,
  ROC = TRUE,
  seed = 1234,
  title = "",
  legendPosition = "bottom",
  labelSize = 2,
  ...
)

## S4 method for signature 'AnalysisData'
plotSupervisedRF(
  x,
  cls = "class",
  rf = list(),
  label = NULL,
  shape = FALSE,
  ellipses = TRUE,
  ROC = TRUE,
  seed = 1234,
  title = "",
  legendPosition = "bottom",
  labelSize = 2
)

## S4 method for signature 'Analysis'
plotSupervisedRF(
  x,
  cls = "class",
  rf = list(),
  label = NULL,
  shape = FALSE,
  ellipses = TRUE,
  ROC = TRUE,
  seed = 1234,
  title = "",
  legendPosition = "bottom",
  labelSize = 2,
  type = c("pre-treated", "raw")
)

Arguments

x	object of class AnalysisData or Analysis containing analysis results
cls	information column to use for sample classes
rf	list of additional parameters to pass to randomForest
label	information column to use for sample labels. Set to NULL for no labels.
shape	TRUE/FALSE use shape aesthetic for plot points. Defaults to TRUE when the number of classes is greater than 12
ellipses	TRUE/FALSE, plot multivariate normal distribution 95% confidence ellipses for each class
ROC	should receiver-operator characteristics be plotted?
seed	random number seed
title	plot title
legendPosition	legend position to pass to legend.position argument of ggplot2::theme. Set to "none" to remove legend.
labelSize	label size. Ignored if label is NULL
...	arguments to pass to the appropriate method
type	raw or pre-treated data to plot

Examples

library(metaboData)

d <- analysisData(abr1$neg[,200:300],abr1$fact)

## Supervised random forest MDS plot
plotSupervisedRF(d,cls = 'day')

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