[R-sig-ME] [R] linear mixed model required for the U.S. FDA

[Helmut Schütz]

Dear Thierry,

Thierry Onkelinx wrote on 2019-08-19 14:13:
> Please do check all parameters. I recall that SAS and R use a 
> different style of dummy coding factor variable.

I’m not a SASian. Hence, I can’t tell. In R factors specified as 
characters are handled in lexical order and internally coded as integers 
starting with 1.

> If that is not the case, please write the full equation for the SAS 
> model. The language of mathematics is the best way to clearly describe 
> a model and thus to compare models.

Agree.
Unfortunately the SAS code is all the FDA gives. I think everything is 
clear except the covariance structure given by FA0(2).
See 
https://support.sas.com/documentation/cdl/en/statug/63033/HTML/default/viewer.htm#statug_mixed_sect019.htm#statug.mixed.mixedrandomtypefa0 
Table 56.13
Sorry, beyond me.

Best regards,
Helmut

-- 
Ing. Helmut Schütz
BEBAC – Consultancy Services for
Bioequivalence and Bioavailability Studies
Neubaugasse 36/11
1070 Vienna, Austria
W https://bebac.at/
C https://bebac.at/Contact.htm
F https://forum.bebac.at/