I am using sas 9.3 version. In the out model data set from proc pls procedure, what does the letters 'WB', 'PQ' and 'V' mean?
I am double checking is W and B are a combination of Weights and Beta coefficients. V is the proportion of variance. What does P and Q refer to? Thanks.
Hi @variety,
Checking some internal docs, I see that PROC PLS OUTMODEL= is undocumented since SAS version 7. As such, it's probably best to not rely on these for production code.
I suspect the same values can be gleaned from ODS OUTPUT, and they might have clearer descriptions. Ron Cody has a good tip on ODS OUTPUT here. And here's something more stat-oriented from @Rick_SAS.
All of the ODS table names for PLS are documented here.
Chris
Thank you for your response @ChrisHemedinger.
My intention in asking the question is: I am curious as to find out what it means.
I personally use OUTPUT= statement to save my results.
Best Regards.
@variety wrote:
Thank you for your response @ChrisHemedinger.
My intention in asking the question is: I am curious as to find out what it means.
I personally use OUTPUT= statement to save my results.
Both Chris and I are referring to ODS OUTPUT which is not the same as the OUTPUT statement. I gave a link in your other thread to documentation on using ODS OUTPUT with PROC PLS, and Chris has done so as well.
As far as what these numbers mean, you need to read some introductory material on Partial Least Squares.
@variety wrote:
I am using sas 9.3 version. In the out model data set from proc pls procedure, what does the letters 'WB', 'PQ' and 'V' mean?
I am double checking is W and B are a combination of Weights and Beta coefficients. V is the proportion of variance. What does P and Q refer to? Thanks.
May I make a suggestion that will improve everyone's efficiency of using the forum?
You don't need to ask the exact same question in two different threads. And if you ask the question in one thread and the reply is to show your code, then if you ask the question again in a different thread ... show your code!!! (But really, don't ask a 2nd time)
But anyway, I gave essentially exact same answer as @ChrisHemedinger in your other thread.
SAS Innovate 2025 is scheduled for May 6-9 in Orlando, FL. Sign up to be first to learn about the agenda and registration!
Learn the difference between classical and Bayesian statistical approaches and see a few PROC examples to perform Bayesian analysis in this video.
Find more tutorials on the SAS Users YouTube channel.
Ready to level-up your skills? Choose your own adventure.