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
- /
- SAS/GRAPH and ODS Graphics
- /
- Correlation matrix of Proc Corr in SAS

Topic Options

- Subscribe to 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
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

05-01-2013 04:33 PM

Hi,

I have two, probably small, problems with the matrix output in proc corr:

1)

Using Pearson correlation coefficient, does anyone know how I can get the level of significance for my matrix-output using proc corr to appear as asterisks (*,** and *** for p<0.1,p<0.05 and p<0.01 respectively) on the correlation coefficient value?

2)

How do I get the matrix to show only half the matrix (since they are identical row/col vs. col/row)?

If it is possible, with the diagonal line of 1:s (representing correlation of x1 on x1 = 1 etc) to either show a histogram of that particular variable or

a scatter plot with another variable not in the matrix (I am thinking of the residuals of y).

I have searched online for hours without finding anything. Help is much appreciated!

Best regards,

/H

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

Posted in reply to Hank

05-02-2013 10:23 AM

Because you want to modify the legend and show only half the matrix, I think you will need to use the GTL to hand-code this layout. To get started, look at the examples (and macro code) that Sanjay provides in this blog post: Compact Scatter Plot Matrix - Graphically Speaking

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

Posted in reply to Rick_SAS

05-09-2013 06:53 AM

Hi,

Thanks for your link to more information, I have looked at it and seems to be a good step in the right direction.

Sorry for the late replay!

Regards,

Hannes