algorithms.pca¶

Module: `algorithms.pca`¶

Step 1 test based on largest singular vector.

This is the test described in Kac Rice for \(X=I\) and the penalty being the nuclear norm

\[{\cal P}(eta) = \sim_{i=1}^{ ext{min(n,p)}} \sigma_i(eta)\]

for \(eta \in \mathbb{R}^{n imes p}\).

selectinf.algorithms.pca.pvalue(X, sigma=1, nsim=5000)[source]¶