Date:         Mon, 4 Jun 2001 10:18:21 -0700
Reply-To:     Paul von Hippel -- Ohio State <pvh@CCRMA.STANFORD.EDU>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         Paul von Hippel -- Ohio State <pvh@CCRMA.STANFORD.EDU>
Subject:      constraining logistic slopes
Content-Type: TEXT/PLAIN; charset=US-ASCII

I'm fitting the following logistic regression model

        log odds (Y) = b_0 + b_1*X_1 + b_2*X_2 + ... + b_k*X_k

and would like to constrain the parameters so that

        b_i >= 0 for all i>=2
                and
        b_1 + b_2 + ... + b_k = 0

Any ideas on how to do this in SAS?

Many thanks,
Paul von Hippel