cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 
JMP is taking Discovery online, April 16 and 18. Register today and join us for interactive sessions featuring popular presentation topics, networking, and discussions with the experts.
Choose Language Hide Translation Bar
Irene
Level I

Tip: confidence ellipses - Princ Comp biplot

Dear JMP community

 

I have runned a Principal Component Analysis. I would like to use the biplot generated but adding a density elipse. This is order to see how my variables distribute regards the data and how powerful is this association.

 

Hope anybody can help on this.

 

Thank you very much in advance

Kind regards to the community

Irene   

2 REPLIES 2
gzmorgan0
Super User (Alumni)

Re: Tip: confidence ellipses - Princ Comp biplot

Irene,

 

I only know of 2 ways to do this, maybe others have a better suggestion. My methods take some editing or scripting. Below is PrinComp1 vs. PrinComp2 for the Iris.jmp data table. For this example I decided to make nonparametric density plots for each group (top left corner of biplot)

Steps:

  1. Save the intended PrinComps. This example saved 3.
  2. Select Fit Y by X. Plot  Y( PrinComp2 ) and  X( PrinComp1 ). Select Group By  Species. Select Nopar Density.
  3. Right click on the framebox of this bivariate, select copy Frame Contents
  4. Right click on the framebox for the corresponding biplot; select paste Frame Contents
  5. Right click on that same biplot framebox and select customize. One at a time select each arrow, change the color to black and 2 wide. Then use the down arrow button to move taht arrow to be the last item in the list. repeat this for each arrow. 
  6. Select OK 

 

Another alternative is to create a Scatterplot Matrix plot of the saved PrinComp. Select GroupBy and the the type of density ellipse you want.  You would have to write a graphics script to create the arrows from the loadings

image.png

Irene
Level I

Re: Tip: confidence ellipses - Princ Comp biplot

Hi!

 

Thank you very much for your reply and for the awesome explanation! I have created very nice graphs! 

Thanks a lot!

Regards

Irene