Date:         Wed, 19 Mar 2008 02:04:40 -0700
Reply-To:     chichibibi <bilbo.vador@HOTMAIL.FR>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         chichibibi <bilbo.vador@HOTMAIL.FR>
Organization: http://groups.google.com
Subject:      Re: the PHREG procedure
Comments: To: sas-l@uga.edu
Content-Type: text/plain; charset=ISO-8859-1

On 18 mar, 18:55, A Sanders <asand...@ufl.edu> wrote:
> On Mar 18, 1:08 pm, chichibibi <bilbo.va...@hotmail.fr> wrote:
>
Thanks a lot Ashley for your help!!
I think now it's gonna be ok with this syntax: "model days*censor(0) =
age drug agedrug;" you gave to me.
have a good day and see you later

Charlotte.


>
>
>
>
> > Hi everybody!
>
> > I'm working on survival data and I would like to test interaction
> > between 2 covariates!
> > but in the model statement of PHREG procédure, I'm not able to write
> > the correct syntax for the interaction.
> > in SAS help, I found the HIERARCHY option which seems to be able to
> > take into account interaction term but it's a new option and probably
> > usable with TPHREG procedure. Unfortunately, I don't know this new
> > procedure!
> > So, does anybody can help me to write the correct syntax for testing
> > interaction?
>
> > Thanks a lot,
>
> > Charlotte
>
> Charlotte,
>
> You can't specify an interaction in the model as you do with other
> procs (e.g. GLM). You can create the interaction term with a separate
> statement within PHREG and then include that term in the model.
>
> proc phreg;
> model days*censor(0) = age drug agedrug;
> agedrug = age*drug;
> quit;
>
> Ashley- Masquer le texte des messages précédents -
>
> - Afficher le texte des messages précédents -