Influence of the input variables on the principal components (PCA)


I have not been using KNIME for long and I am just learning the ropes. I hope my question is not stupid. I have edited a data set and want to reduce it to two dimensions once using PCA once using t-SNE. This also works. Now I want to know which input variables have how much influence on my principal components, in order to display this in a ranking at the end. Unfortunately, I don’t know which nodes are best suited for this. Rank Correlation makes no sense to me and with Linear Correlation I get the error message that a column has the wrong name.

I hope you can help me!


Hi @lucam94 ,
Welcome to the KNIME community!
Here are some of the related workflows to PCA and t-SNE that might help you

Let us know if these workflows helped you. You could also explore other workflows related to this topic on the KNIME Community Hub



See if this helps. Same data set as the first link @sanket_2012 sent you. I have no idea what your data looks like. I’m not a trained statistician, so I’m not getting in a discussion about which correlation method is better.


This topic was automatically closed 90 days after the last reply. New replies are no longer allowed.