Date: Wed, 2 May 2007 08:59:56 -0400
Reply-To: Raghu Venkat <raghustays@GMAIL.COM>
Sender: "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From: Raghu Venkat <raghustays@GMAIL.COM>
Subject: Re: Which SAS procedure to use
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
try to first see whats in the data.
a correlation matrix using proc corr might be a good starting point to see
if any multicollinearity
issues exist. Then see if you can combine some of those variables and reduce
it some other variable that might make sense.
i dont think anyone here will have a ready made solution to your question.
This sounds like one of the class projects :)
On 5/2/07, Kunal Kelkar <email@example.com> wrote:
> Hi All,
> I am having a SAS dataset with close to 640
> variables out of which 430 are of categorical type.I am trying to fit
> regression model using this data. My response variable is also of
> categorical type.Many variables are having lot of missing values.I need
> help to find out how I can reduce the number of variables and how can tackle
> the null values.Also which SAS procedure will be helpful for this. If
> possible if anyone can provide some example that will be great.
> Thanks in advance.
> Check Out the new free AIM(R) Mail -- 2 GB of storage and industry-leading
> spam and email virus protection.