Emax model fit

[hui zhang]

Hi,  Could we fit the PK/PD Emax model using stats models. Or we can add sigmoid emax model to the api. Many thanks

[josef...@gmail.com]

Can you be more specific about what the model and parameters are?

From what I understand looking briefly at some online resources, which, based on only superficial skimming, are not very explicit about what parameters are estimated:

My guess is that statsmodels cannot estimate (most of) those models because we don't have a general nonlinear model estimator.

Our main models that are linear in parameters are based on link functions with linear predictor inside, that is

E(y | x) = f(x * beta)

where f maps into a fixed support, R, R+ or [0,1] interval.

For example sigmoid have known supports limits [0, 1]   (or [0, n], n known) for fractional or binomial data.

All other models assume that the mean function is linear in parameters.

The best python package for nonlinear least-squares with user provided nonlinear mean function is lmfit.

Josef

On Fri, Apr 26, 2024 at 8:08 AM hui zhang <zhui...@gmail.com> wrote:

Hi,  Could we fit the PK/PD Emax model using stats models. Or we can add sigmoid emax model to the api. Many thanks

--
You received this message because you are subscribed to the Google Groups "pystatsmodels" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pystatsmodel...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/pystatsmodels/929ca18e-fbda-49eb-801d-8808376f7d7an%40googlegroups.com.