sas >> Proc GLM and Estimate statement

by fongleung » Tue, 25 Nov 2008 04:30:58 GMT

Hi,

I am trying to save the estimate and standard error from the estimate
statement. I know there is an output statement which allows me to save
R, LCL, and UCL etc. But this does not allow me to save the number
generated by the estimate statement. All I need is to save the
estimated LS Means and Std Error for further calculation and
reporting.

Does anyone know how to do?

Thanks.


sas >> Proc GLM and Estimate statement

by lewjord » Tue, 25 Nov 2008 05:02:50 GMT


ODS output estimates, should work.

---- Original message ----



Similar Threads

1. estimate Y-hat in PROC GLM

Dear SAS-L experts,

I am trying to get estimate for Y-hat for each class of the interaction term
in GLM, I couldn't get the estimate statement  right and need your help.

Here is my model:

PROC GLM;
CLASS ETHNGRP;
MODEL CPD=GROUP ETHNGRP GROUP*ETHNGRP/SOLUTION
ESTIMATE 'GROUP1 AND WHITE' GROUP*ETHNGRP ????;
ESTIMATE 'GROUP2 AND WHITE' GROUP*ETHNGRP ????
RUN;

I really don't know how the group*ethngrp matrix formed here, would you
please provide a solution here?

Thanks in advance.

Sophia

2. How to write ESTIMATE in PROC GLM

3. PROC GLM parameter estimates

Hi

Here's an embarrassingly simple question.  If I want parameter estimates
from PROC REG, I can save them in an OUTEST data set.  How do I save
parameter estimates from PROC GLM?  Sorry, I couldn't find it from the
documentation.  Thanks!

--  TMK  --
"The Macro Klutz"

4. Contrasts/estimates with proc GLM

5. PROC GLM: Contrast/Estimate

6. PRESS when using WEIGHT statement in PROC GLM

7. contrast statements and proc GLM code problem

person is not on SAS L

is there any good references on proc GLM and how to set up contrast statements?
I've sent info on archives, but that and the manual is not real clear.

thanks.
-----Original Message-----
From: Kraft, Ottalee (UMC-Student) 
Sent: Monday, August 11, 2003 5:46 PM
Subject: SAS Contrast Statements


After working with Ron today I have solved some of my SAS problems but I am still stuck on the contrast statements to get the comparisons I want.  My class variable, Type represents 4 conditions.  I want to be able to compare Type 1 to Types 2, 3, and 4 combined; to compare Type 2 to Types 3 and 4 combined; to compare Type 3 to Type 4.   Ron initially helped me set up some IF statements to break my data by type and then run the proc glm.  However, as I understand statistical comparisions, this partitioning of my data will result in incorrect analyses.  I have a couple of questions I need help with:
     1. How do I set up the contrast statements in SAS to do the comparisions I need?
     2. Is the repeated measures ANOVA or the MANOVA preferable for any reason?  I seem to get comparable results.
     3. I will need to add some additional covariates to these models ( age,  gender, and a couple of yes/no type responses).  Does this change your responses to questions 1 & 2?

The code I have set up so far is:
____________________________
**REPEATED MEASURES ANOVA**
____________________________
proc glm data=diss;
class Type;
model RatingMHCX
      RatingMHCY
      RatingMHCZ = Type / nouni;
repeated rating 3 contrast (1) / summary;
lsmeans Type / stderr pdiff cov out=adjmeans;
title1 'results of repeated measures glm';
quit;
run;

__________________________
**         MANOVA       **
__________________________
proc glm data=diss;
class Type;
model RatingMHCX
      RatingMHCY
      RatingMHCZ = Type;
means Type / Tukey;
means Type;
manova H = Type;
title1 'results of glm -- manova';
quit;
run;


**********************************************************************
This electronic communication is from the Missouri Department of Health and Senior Services and is confidential, privileged and intended only for the use of the recipient named above.  If you are not the intended recipient or the employee or agent responsible for delivering this information to the intended recipient, unauthorized disclosure, copying, distribution or use of the contents of this transmission is strictly prohibited.  If you have received this message in error, please notify the sender immediately at the following email address:  XXXX@XXXXX.COM  or by calling (573) 751-6400..  Thank you.
**********************************************************************

8. How to write estimate statement (Proc Mixed) for interaction