Exercise

# PCA with ade4

Alright! Now that you've got some real hands-on experience with `FactoMineR`

, let's have a look at `ade4`

, a well-known and well-maintained R package with a large number of numerical methods for building and handling PCA models.

`dudi.pca()`

is the main function that implements PCA for `ade4`

and by default, it is interactive: It lets the user insert the number of retained dimensions. For suppressing the interactive mode and inserting the number of axes within the `dudi.pca()`

function, you need to set the `scannf`

argument to `FALSE`

and then use the `nf`

argument for setting the number of axes to retain. So, let's put `ade4`

into practice and compare it with `FactoMineR`

.

Instructions

**100 XP**

- Run
`dudi.pca()`

on the numeric variables of`cars`

with the`ade4`

package and, in a**non-interactive mode**, extract the first**4**principal components. - In the next two lines of code, you are asked to explore with
`summary()`

the resulting object`cars_pca`

and`pca_output_ten_v`

, the PCA model object that you built with`FactoMineR`

earlier that is available in your workspace. Spend a few minutes in comparing the difference in the output of the two summaries.