Compute variance explained by each Principal Component
We use the formula below to compute the total variance experienced by each PC.
Syntax: pca$sdev^2 / sum(pca$sdev^2)
Code:
R
# drop the species column as its character typenum_iris = subset(iris, select = -c(Species) )# compute pcapca <- prcomp(num_iris, scale = TRUE)# compute total variancevariance = pca$sdev^2 / sum(pca$sdev^2)variance |
Output:
[1] 0.729624454 0.228507618 0.036689219 0.005178709
Example 1: Plotting Scree plot with Line plot
R
library(ggplot2)# drop the species column as its character typenum_iris = subset(iris, select = -c(Species) )# compute pcapca <- prcomp(num_iris, scale = TRUE)# compute total variancevariance = pca $sdev^2 / sum(pca $sdev^2)# Scree plotqplot(c(1:4), variance) + geom_line() + geom_point(size=4)+ xlab("Principal Component") + ylab("Variance Explained") + ggtitle("Scree Plot") + ylim(0, 1) |
Output:
Example2: Plotting Scree plot with barplot
R
library(ggplot2)# drop the species column as its character typenum_iris = subset(iris, select = -c(Species) )# compute pcapca <- prcomp(num_iris, scale = TRUE)# compute total variancevariance = pca $sdev^2 / sum(pca $sdev^2)# Scree plotqplot(c(1:4), variance) + geom_col()+ xlab("Principal Component") + ylab("Variance Explained") + ggtitle("Scree Plot") + ylim(0, 1) |
Output:
How To Make Scree Plot in R with ggplot2
In this article, we are going to see how can we plot a Scree plot in R Programming Language with ggplot2.