1
$\begingroup$

I've been trying to find a way to fit a generalized nonlinear model in R, with little success. The stochastic component would be the following:

$$Y \sim binomial(n, pi)$$

The parameters being b1 and $\gamma$

The systematic component is the following:

$$ pi = \frac{e^{b1*(conc - \gamma)}}{1+e^{b1*(conc - \gamma)}} $$

In SAS this model can be fit using the following code:

data one;
  do j=1 to 6; 
    input conc n y @@; conc=conc/10;
    output; 
  end; datalines;
501 48 48 407 50 47 302 49 47 204 48 34 100 48 18 51 49 16
;

proc nlmixed data=one;
  parms b1=1 gam=10;
  ex=exp(b1*(conc-gam));
  pi=ex/(1+ex);
  model y~binomial(n,pi);
run;

Any ideas on how to fit this in R?

$\endgroup$
  • $\begingroup$ The package gnm is explicitly for generalized nonlinear models $\endgroup$ – Glen_b Sep 24 '15 at 18:07
  • $\begingroup$ @Ian have you looked at the locfit package with implements local (i.e., nonlinear, nonparametric) GLM? $\endgroup$ – Antoine Sep 24 '15 at 18:53

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.