Search code examples
rpca

PCA plot, dimensions plot

I have found good tutorial about PCA One question is not clear me currently. I want to know, how is it possible to visualize not only Dim1 vs Dim2 results, but actually all possible pairs of dimensions (for example Dim3 vs Dim4)

Solution

  • You can visualize the other dimensions by changing the axes(1,2) argument to e.g. axes(3,4):

    fviz_pca_ind(X, axes = c(3, 4), geom = c("point", "text"),
   label = "all", invisible = "none", labelsize = 4) 
# (...)

    Side note: The first couple of principal components often contain almost all variation in the dataset. The last principal components are generally very uninteresting, because the variation in the dataset has been "summarized" by the preceding principal components, and therefore only contains residue information (noise).