Date:         Fri, 16 Oct 1998 11:19:47 -0700
Reply-To:     Siyan Wang <siyanwan@RCF.USC.EDU>
Sender:       "SAS(r) Discussion" <SAS-L@UGA.CC.UGA.EDU>
From:         Siyan Wang <siyanwan@RCF.USC.EDU>
Subject:      Questions about SAS/IML
Comments: cc: sas-l@vm.marist.edu, sas-1@vtvml.cc.vt.edu
Content-Type: TEXT/PLAIN; charset=US-ASCII

I get some problem using SAS/IML. so I would like to ask for your help.
 
1. I have a SAS data set, which consist of 18 variables and 360
observations. I want to use this data set to creat a matrix consisting of
17 variables (row) over all 360 observations (colomn). Also, I want to use
the variable left to creat a colomn vector.
 
2. I used PROC CORR procedure to calculate a 20 by 20 covariance matrix,
The statement used is
  PROC CORR DATA=IR.RES_N NOMISSING NOSIMPLE NOCORR OUTP=IR.COV_N COV;
How can I read these covariance into a matrix by IML?
 
3. How can I do Kroneker product?
For example, the Kroneker product of a 3 by 3 identity matrix and a 4 by 4
matrix Q is
               / Q     \
               |   Q   |
               \     Q /
i.e. the Kroneker product is a 12 by 12 block diagonal with matrix Q as
the diagnol blocks and 0 for all other elements.
I could not find the proper operator or function in the manul.
 
Thank you very much.
 
Siyan