Bioequivalence and Bioavailability Forum • Let's skip the fancy nesting syntax!

Let's skip the fancy nesting syntax! [🇷 for BE/BA]

posted by ElMaestro – Denmark, 2009-08-28 00:46 (6129 d 13:25 ago) – Posting: # 4130
Views: 15,800

Hi Yjlee,

❝ What if we keep same data matrix as before (e.g., as in bear), but change the model with "lnAUC ~ (1|Subj:Seq) + Seq + Per + Trt"? I haven't tried it yet. I read this from (...) It seems that we will get normal "Subj: Seq" in anova table in this case.

I do not know how R will handle that, sorry. But I don't personally have a preference for it. I clearly would prefer the simple lnAUC~Per+Trt+Seq+Subj in combo with uniquely coded subjects.
If you really insist on the ":" in the model syntax (please tell me why!) and you need to make sure the output for the nesting is never specified as "Seq:Subj" then I guess you could also fiddle with some aspects of R's data garbling. What in practice this implies, I could only tell if I read the source code. That is not likely to happen any time soon, though.

EM.

Complete thread:

Let's skip the fancy nesting syntax! ElMaestro 2009-08-26 20:54
- Let's skip the fancy nesting syntax! yjlee168 2009-08-26 22:29
  - Let's skip the fancy nesting syntax! ElMaestro 2009-08-26 22:46
    - Let's skip the fancy nesting syntax! yjlee168 2009-08-27 20:54
      - Let's skip the fancy nesting syntax! ElMaestro 2009-08-27 22:14
- Let's skip the fancy nesting syntax! yjlee168 2009-08-27 22:26
  - Let's skip the fancy nesting syntax!ElMaestro 2009-08-27 22:46
  - Simple solution ElMaestro 2009-08-27 23:53
    - Simple solution yjlee168 2009-09-07 01:53
      - Type III SS again ElMaestro 2009-09-07 15:47
        
        Type III SS again yjlee168 2009-09-07 18:08
        
        Type III SS again ElMaestro 2009-09-07 19:53