How can I create a zero-truncated binomial distribution pf?

2 views (last 30 days)
Hi all,
The page http://www.mathworks.com/help/stats/examples/fitting-custom-univariate-distributions.html shows how to make a zero-truncated Poisson probability function for MLE fitting. Does anyone know how to do the same thing for a binomial distribution?
I have the following code:
function paramests = fitBinoMixture(hist)
N = 10;
pdf_binomixture = @(x, t, p1, p2)...
t*binopdf(x,N,p1) + (1-t)*binopdf(x,N,p2);
%Parameters
t_start = 0.5;
p1_start = 0.03;
p2_start = 0.5;
start = [t_start p1_start p2_start];
lb = [0 0 0];
ub = [1 1 1];
%Rum MLE
paramests = mle(hist, 'pdf',pdf_binomixture, 'start',start, 'lower',lb, 'upper',ub);
This works on simulated data, but not on my experimental data in which zero-occurrence events are not observed.
Thanks for any advice! Dan

Answers (0)

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!