PCA with FactoMineR

As you saw in the video, FactoMineR is a very useful package, rich in functionality, that implements a number of dimensionality reduction methods. Its function for doing PCA is PCA() - easy to remember! Recall that PCA(), by default, generates 2 graphs and extracts the first 5 PCs. You can use the ncp argument to manually set the number of dimensions to keep.

You can also use the summary() function to get a quick overview of the indices of the first three principal components. Moreover, for extracting summaries of some of the rows in a dataset, you can specify the nbelements argument. You'll have a chance to practice all of these and more in this exercise!

As in the previous lesson, the cars dataset is available in your workspace.

Run PCA() on the 10 numeric variables (9 to 19) of cars and extract 4 dimensions. To set the number of dimensions, you need to use the ncp argument of PCA(). The graph argument is set to F to suppress the default graphical output PCA().

Take Hint (-7 XP)

Exercise

PCA with FactoMineR

Instructions 1/4