turn on suggestions

Auto-suggest helps you quickly narrow down your search results by suggesting possible matches as you type.

Showing results for

Find a Community

- Home
- /
- SAS Programming
- /
- General Programming
- /
- Correlation Matrix Problem

Topic Options

- RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page

- Mark as New
- Bookmark
- Subscribe
- RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

08-18-2014 06:09 AM

I have a problem with Correlation matrix output.

As in the attached excel, this is my correlation matrix output dataset.

Please igore the yellow highlighted, as it is the replica and the red part, it is the correlation of same variables.

My interest is in the below the non-highlighted area.

My requirement,

B*A =.98

C*A=.98

Then it is obvious that B*C=.98. So, if that is the case I dont want this to be populated as .98, instead it should be populated as NULL or someother numeric value.

Basically i dont want the redundant correlation. Like this I have a 300*300 matrix. I think this is a complex problem, but your help is much appreciated.

- Mark as New
- Bookmark
- Subscribe
- RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

Posted in reply to SOORISAS

08-18-2014 09:13 AM

This is not redundant information.

You can calculate upper and lower bounds for corr(B,C) from corr(A,B) and corr(A,C), but not the exact value.

- Mark as New
- Bookmark
- Subscribe
- RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

Posted in reply to SOORISAS

08-18-2014 09:16 AM

Then it is obvious that B*C=.98.

I don't think this is a true statement, I'm pretty sure that the correlation B*C can be numbers other than 0.98, and so I don't see how to proceed with the rest of your request, nor do I think it reasonable at this time to continue with the rest of your request.

--

Paige Miller

Paige Miller

- Mark as New
- Bookmark
- Subscribe
- RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

Posted in reply to SOORISAS

08-19-2014 10:56 AM

If B=C, then corr(B,C)=1.

- Mark as New
- Bookmark
- Subscribe
- RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

Posted in reply to SOORISAS

08-19-2014 12:18 PM

Hi,

Why not eliminate redundant variables first using some data reduction technique? This will automattically eliminate redundant correlation.