Help!

[Metz]

Hi! I have a question: how can i 'translate' this RStudio command

     > fm9.1<-gls(visual~-1+visual0+time.f+treat.f:time.f, weights=varIdent(form=~1|time.f),data=armd)

     > summary(fm9.1)

in SAS code?

I can't solve this problem. Thanks for your help!

[Steelers_In_DC · Accepted solution]

I do not know Rstudio but if you give a sample of data you have and what you are trying to do I can try to help.

[RW9]

Hi,

With regards to the Linear Modelling GLS procedure, you can have a look at:

http://support.sas.com/documentation/cdl/en/statugregression/61752/PDF/default/statugregression.pdf

Or post directly into the statistics section on here about Linear Models.

With regards to the summary command, there is proc summary, proc means, proc freq.  All have help pages on line.  Should be mainly putting your variables in the correct order.

[Metz]

Thanks! But my problem is that hat i want to do a linear regression with heterogeneus variance, because my 'w_i' is not equal 1 but is equal to the variance of the first level of a model's factor. How can i write this?

[RW9]

I would suggest posting in the Statistical Procedures section of this help with your question, I am not a statistician sorry.

[Ksharp]

You mean Weighted Least Square ? Post it at Stat Forum.  I guess proc glm can do that.

[Metz]

Yes, i mean that! Ok i try there. Thank you all