Why the 95%CI and CV% of Theta could be calculated in FOCE ELS engine.

Dear all,

I’m working on a population model based on a rich dataset. I get initial estimates from the Naive Pool model. But after I accepted all fixed and randome effects and changed the engine to FOCE ELS(or L-B), the model could run normally. But when I check Theta worksheet, I could not get the 95%CI and CV% of all parameters. The attached project file is for your infoemation. The object name in this file is “FOCE additive Two cpt Model Tlag”.Could you please help me check If there are wrong settings in my object? I will Highly appreciate if you can clear it.

Thank you very much & best regards.

Fu Yangyang

20190121-3.phxproj (3.54 MB)

Hi Yangyang, this was a but tricky to get SEs for.

What I did was try QRPEM first that got me SEs, then I tried using those final parameter esitmates but still no SEs/.

then I swapped to Fisher and Forward diff for SE calc and got the results attached.

I woudl comment that there may not really be enough data to support 2 compartments, and the omega shrinkage for CL2 and V2 is also a little high (0.4090162 and 0.50924805)

. there is a fair bit of noise inthe absorption and it is only a couple of points on some profiles that suggest 2com so you may want to reduce to 1 compartmnet and perhaps even try other residual error models e.g. Poisson or multiplicative.

Simon.20190121.phxproj (5.56 MB)

Dear Simon,

Thank you.

I have another qustion. I use the project file you run, and copy the object “QRPEM additive Two cpt Model Tlag”, then I accepted the final parameters and rerun the QRPEM object. it would prompt the error message like figure 1. Could you help me check it? Thank you

In addition. In my data, there are ten covariates for this population model. and some of them are relevant, such as group of weight, height and BMI. The detail correlationship could refer the attached plot. On this occasion, how can I select the covariate in the stepwise? Could you please give me some advices? Thank you very much.

Sincerely,

Fu Yangyang