3 replies to this topic

[Gilles TUFFAL]

Hi, 

 

In order to foster Phoenix NLME use, and despite the possibility to use comparator tool, is there a way to recover Omega/RSE  and Sigma/RSE in an easier way?

 

The joined table shows how we use to summarize data (POPPK summary).

 

Is there a way to recover Omega², their SE and RSE (calculate) in an easier way than in 2D tables (Omega, Omegastderr) and grab for diagonal (or in OOD for omegablocks).

May the tabulated output gathering all exits in any output ?

And/or would it be possible creating this output by means of R shell or Phoenix wokflow in a generic way and ready usable as for Theta table.

 

By the time, one could clarify about CV vs RSE, some data presented as CV  being indeed RSE.

 

Many thanks.

 

Gilles

Attached Files

•  POPPK summary.jpg   86.24KB   1 downloads

Edited by Gilles TUFFAL, 29 April 2021 - 09:39 AM.

[smouksassi1]

Hi Gilles,

did you use or see RsNLME you can do everything from R command line but I am not aware of an automatic way to generate this kind of table including transforming the standard error of Omega

Phoenix gives you omega variance and the SE RSE on the variance scale

 

if you want to report on 

squaroot of omega or on sqrt (exp(omega) -1) it is up to you to compute the correct se using the delta method as the omega model can be anything 

of course if you wrok in R you can program the whole thing and parse which parameter had exp eta and which had an additive etc.

[kniefort]

Hi Gilles,

 

You may wish to take a look at using Pirana with RsNLME.  This type of summary is done automatically for you.  You can also push output summaries into a word document (or html/pdf/LaTeX).  See attached examples.

 

If you're interested, we have an upcoming webinar that will highlight some of the reporting workflow support tools we're working on for our Pirana/RsNLME tools.

 

https://www.certara.com/webinar/upcoming-2021-certara-university-express-webinars/

 

 

 

Attached Files

•  PiranaRsNLMEScan.png   329.26KB   1 downloads
•  pirana_run_summary.docx   16.55KB   10 downloads

[Gilles TUFFAL]

Thank you both, 

 I was just wondering if I missed something in Phoenix NLME (alone). A good reason to move to Pirana/RsNLME. On going a fast track training on R (basics essentially) to better exploit RsNLME possibilities.

The Pirana outputs is just what I expected. Report ready !

Best.

Gilles