Bioequivalence and Bioavailability Forum

Optimizer settings [Software]

posted by Helmut  – Vienna, Austria, 2014-10-09 16:08 (3859 d 22:13 ago) – Posting: # 13680
Views: 38,790

Dear Yung-jin,

❝ Have you ever tried to add/change options of MAXITER (default is only 50; may consider to increase this value) and / or MAXFUNC (default is 150) with PROC MIXED when getting convergence failure for ABE?

I’m not gifted with
In Phoenix the default maximum iterations are also 50. Increasing doesn’t help, since I never have seen more than seven for datasets ending up with a warning:

Newton's algorithm converged with a modified Hessian, indicating that the model may be over-specified. The output is suspect. The algorithm con­verged, but not at a local minimum. It may be in a trough, at a saddle point, or on a flat surface.

❝ No starting value can be set up with either PROC MIXED of SAS or lme() from package nlme of R.

Datasets failing to converge in Phoenix:

Initial variance matrix is not positive definite. This indicates an invalid starting point for the algorithm.

However, initial values in Phoenix REML are obtained by the Method of Mo­ments (MoM). It is not possible to set them “manually”.

—
Dif-tor heh smusma 🖖🏼 Довге життя Україна! 
Helmut Schütz


The quality of responses received is directly proportional to the quality of the question asked. 🚮
Science Quotes

Complete thread:

• reference dataset elba.romero 2013-07-02 05:08 
  • Public datasets Helmut 2013-07-02 15:44
    • Public datasets yjlee168 2013-07-03 03:10
      • Public datasets elba.romero 2013-07-03 05:19
        • Public datasets yjlee168 2013-07-03 13:11
          • Challenging data sets Helmut 2013-07-03 16:32
            • Fantastic idea? ⇒ Challenging data sets yjlee168 2013-07-04 08:06
              • Fantastic idea! Helmut 2013-07-04 13:18
      • NIST; cross-validation Helmut 2013-07-03 14:28
        • NIST; cross-validation yjlee168 2013-07-03 23:47
          • NIST; cross-validation Helmut 2013-07-04 00:45
    • Public datasets elba.romero 2013-07-03 05:10
  • Eight Reference Datasets (2×2×2 ☑) Helmut 2014-10-06 02:24
    • What??? ElMaestro 2014-10-06 13:20
      • What??? Helmut 2014-10-06 17:08
        • What??? ElMaestro 2014-10-06 17:31
          • What the heck‽ Helmut 2014-10-06 19:05
    • link of electronic supplement material? yjlee168 2014-10-07 09:53
      • link of electronic supplement material? nobody 2014-10-07 10:59
        • Thanks! yjlee168 2014-10-07 11:35
          • Thanks! nobody 2014-10-07 11:44
            • Direct links & hint Helmut 2014-10-07 13:54
      • Replicated designs are relevant, too ElMaestro 2014-10-07 12:36
        • Replicated designs are relevant, too yjlee168 2014-10-07 13:00
          • RSABE: SAS and PHX/WNL only? Helmut 2014-10-07 14:47
            • for the future? yjlee168 2014-10-08 10:05
              • Star Trek: The Next Generation Helmut 2014-10-09 00:26
                • Star Trek: The Next Generation d_labes 2014-10-09 08:09
                • Star Trek: The Next Generation nobody 2014-10-09 08:23
                • Star Trek: The Next Generation yjlee168 2014-10-09 11:06
                  • Some comments ElMaestro 2014-10-09 11:52
                    • Some comments yjlee168 2014-10-09 12:18
                      • Proc MIXED - start value d_labes 2014-10-10 08:35
                        • Proc MIXED - start value ElMaestro 2014-10-10 11:44
                        • Proc MIXED - start value yjlee168 2014-10-11 21:09
                          • Proc MIXED - start value ElMaestro 2014-10-11 21:18
                  • Optimizer settingsHelmut 2014-10-09 14:08
                    • Optimizer settings nobody 2014-10-09 15:32
                    • Optimizer settings yjlee168 2014-10-09 18:14
                      • FDA’s model… Helmut 2014-10-09 18:34
    • Just received a letter… Helmut 2014-11-03 14:54
    • 11 Reference Datasets (2-group parallel ☑) Helmut 2014-12-09 15:51