Do a principal component analysis of all transformed variables simultaneously. Make a graph of the number of components versus the cumulative proportion of explained variation. Repeat this for...

Do a principal component analysis of all transformed variables simultaneously. Make a graph of the number of components versus the cumulative proportion of explained variation. Repeat this for laboratory variables alone.

1. Repeat the overall PCA using sparse principal components. Pay attention to how best to solve for sparse components, e.g., consider the lambda parameter in sPCAgrid.