0

PCA Screeplot

Following the usual methodology I have created a screeplot to assess how many principal component axes are significant. However, the broken stick is greater than the variance for PC3 and PC4 but not PC2.

Is only PC1 significant under the broken stick model or are PC3 and PC4 also significant?

Thanks

steve24
  • 35
  • 3

1 Answers1

0

You are better off asking this on stats SO, but:

The point at which the scree plot curve crosses the broken stick model distribution is considered to indicate the maximum number of components to retain.

Source

AK88
  • 2,946
  • 2
  • 12
  • 31