User Community
cancel
Turn on suggestions
Auto-suggest helps you quickly narrow down your search results by suggesting possible matches as you type.
Showing results for 
Show  only  | Search instead for 
Did you mean: 

JMP Wish List

We want to hear your ideas for improving JMP. Share them here.
Choose Language Hide Translation Bar
0 Kudos

DOE platform improvement suggestions

Status: Needs Info Submitted by on ‎10-22-2019 09:27 AM
2 Comments (2 New)

I would like the ability to get alias effects matrix using JSL in Custom Design.  

  • there is a <<Get X Matrix  (X) that returns the design matrix for the specified model and there is a <<Get Alias Matrix message that gets the aliasing correlations (for model x alias interactions, X`*A)
  • but there is not command to get the alias effects matrix for the design (A)

In Custom Design, Evaluate Design, and Compare Designs, I would like to be able to use JSL get the full correlation matrix that is used to construct the color map on correlations.  So a <<Get Correlations Matrix message would be nice.to add.

 

For Evaluate Design, I would like to be able to see a list of the pairwise correlations of the model x model, model x alias, and alias x alias correlations.  Similar to the pairwise correlations table that can be displayed in Multivariate.  The reason so so that I can sort and see the pairs of effects with the highest confounding.  

 

For Compare Designs, I would like to be able to see the difference in the effects and aliases correlations matrices.  So if I compare 2 designs, and each has a color map on correlations (C1 and C2), I want to be able to see (C1-C2) as a "delta color map".  

2 Comments
Jeff_Perkinson
Community Manager Community Manager
Community Manager
‎02-03-2020 09:53 AM
‎02-03-2020 09:53 AM
Status changed to: Needs Info

Any chance you could split these out into multiple wishes, so we can track them individually?

MathStatChem
Level VII
Friday
Friday

I am pinging this again.  Don't feed the need to separate into separate wishes, as they are all related.  This type of information is really useful to have when you are fitting supersaturated designs or observational data and you want to know which model effects are estimable or not.