pca_plots: Create nice PCA plots.

In ricoderks/Rcpm: General functions for the Center of Proteomics and Metabolomics (LUMC)

Create nice PCA plots.

Description

Generate different plots for a PCA analysis. Score plot (includings Hotelling T2 confidence interval ellipse), loading plot or a summary plot. The summary plot contains a score plot, loadings plot and the summary of fit plot.

Usage

pca_plots(
  data,
  x,
  y,
  type = c("scores", "loadings", "summary"),
  T2 = TRUE,
  hotelling = 0.95,
  colour = NULL,
  shape = NULL,
  size = 3
)

Arguments

data	dataframe containing all information. See details for more information on the structure.
x	Principal component for the x-axis.
y	Principal component for the y-axis.
type	Which plot to generate: scores, loadings or summary plot.
T2	add a Hotteling T2 ellipse to the scores plot. Default is true.
hotelling	Set the Hotelling confidence interval. Default = 0.95.
colour	Set the color of the points for different groups.
shape	Set the shape of the points for different groups.
size	Set the size of the points. Default is 3.

Details

data should be data frame which contains at least 2 variables for plotting the scores or loadings plot. The name of these 2 variables are also used as label for the axes. More variables can be added for coloring or shaping the points. These variables should be factors. See the examples for a simple example.

Value

A ggplot2 plot is returned.

Author(s)

Rico Derks

Examples

# create a dummy frame 
mydata <- data.frame(PC1 = rnorm(40),
                     PC2 = rnorm(40),
                     groups = as.factor(c(rep(1, 20), rep(2, 20))),
                     shapes = as.factor(c(rep(1, 20), rep(2, 20))))
p <- pca_plots(data = mydata, 
               x = PC1,
               y = PC2,
               type = "scores",
               colour = groups,
               shape = shapes)
# or
library(dplyr)
mydata %>%
  pca_plots(x = PC1,
            y = PC2,
            type = "scores",
            colour = groups,
            shape = shapes)

ricoderks/Rcpm documentation built on May 18, 2022, 7:49 a.m.