cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 
Try the Materials Informatics Toolkit, which is designed to easily handle SMILES data. This and other helpful add-ins are available in the JMP® Marketplace
Choose Language Hide Translation Bar
efisch12
Level I

PCA - create loading matrix for each sample, instead of for each variable

Is there a way to view a loading matrix for each sample, showing the variation of each element? I envision a different table for each principal component. 

 

The picture is an example of the table I need. The PCA was run for all of the elements in the table, but for 17 samples. 

 

ih_0-1664835551354.png

 

 

Thank you!! 

 

(I am using the most recent version of JMPpro)

1 ACCEPTED SOLUTION

Accepted Solutions
ih
Super User (Alumni) ih
Super User (Alumni)

Re: PCA - create loading matrix for each sample, instead of for each variable

I think there is some confusion here: loadings would not change for each sample, they are the same for the entire table.  I think you might be looking for the score contributions, and you can see those using the MDMCC platform.  After launching the platform using scores saved from PLS, look for Monitor the Process > Score Plot, and then hover over a point in the scatterplot, or look at the contributions heat map.  Here is an example:

 

 

View solution in original post

2 REPLIES 2
ih
Super User (Alumni) ih
Super User (Alumni)

Re: PCA - create loading matrix for each sample, instead of for each variable

I think there is some confusion here: loadings would not change for each sample, they are the same for the entire table.  I think you might be looking for the score contributions, and you can see those using the MDMCC platform.  After launching the platform using scores saved from PLS, look for Monitor the Process > Score Plot, and then hover over a point in the scatterplot, or look at the contributions heat map.  Here is an example:

 

 

efisch12
Level I

Re: PCA - create loading matrix for each sample, instead of for each variable

That is what I am looking for, thank you very much!