Date:         Tue, 27 Jan 2009 07:57:40 -0800
Reply-To:     creubenassal@GMAIL.COM
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         creubenassal@GMAIL.COM
Organization: http://groups.google.com
Subject:      Re: predicted values
Comments: To: sas-l@uga.edu
Content-Type: text/plain; charset=ISO-8859-1

The variable "prov" has three values: 3, 12, 46 and the input data set
is arranged by this variable. I want three predictions, for prov=3,
prov=12 and prov=46. I have append three rows at bottom of input data
file (r131) with information of "A3" and "prov" variables. The model
estimation is correct but the output give us calcul problems in
predicted values.
Any idea?

On 26 ene, 18:46, lewj...@UGA.EDU (Lewis Jordan) wrote:
> I take it your wanting to make predictions for a new "prov"?
>
> All you should have to do is "append" your prediction data set to your fitting data set "r131". Make sure that you have all of the information PROC MIXED needs to calculate the predicted values, and also that you have a "prov" identifier for the new data, like prov=99999.
>
>
>
> ---- Original message ----
> >Date: Mon, 26 Jan 2009 11:00:55 -0500
> >From: Arthur Tabachneck <art...@NETSCAPE.NET>
> >Subject: Re: predicted values
> >To: SA...@LISTSERV.UGA.EDU
>
> >While I'm not sure what you're asking, have you looked into using proc
> >score?
>
> >Art
> >------
> >On Mon, 26 Jan 2009 07:33:12 -0800, creubenas...@GMAIL.COM wrote:
>
> >>I want to calculate predicted values for data points other than those
> >>observed with PROC MIXED and OUTPRED statement. I have an input data
> >>set with missing values dependent variables, my specification has a
> >>REPEATED statement with the SUBJECT=INTERCEP option, like this:
>
> >>proc mixed data=r131 MMEQ;
> >>class prov;
> >>model b13=a3 /s ddfm=kr outp=r131pm;
> >>random prov / s;
> >>repeated / type=lin(1) ldata=ldata1 subject=intercept;
> >>run;
>
> >>Where have I to put the individuals with information in the
> >>independent variables and with missing dependent variable in my input
> >>data set in order that the previous senetences calculate the
> >>corresponding predicted values, standard errors and related
> >>quantities?- Ocultar texto de la cita -
>
> - Mostrar texto de la cita -