Inverse CDF for Generalized Linear Model (GLM) with Binomial Distribution

I am using glmfit to create a GLM that has a binomial distribution and logit link function.

The inputs to glmfit are a range of voltages (predictors) and the binary response at each of those voltages (i.e. worked or didnt work)

I am able to calculate the probability of working at a specific voltage by using glmval. However, I would also like to do the inverse of that and calculate the voltage at a given probability. For example, at what voltage is there a 50% probability of working?

Any help is much appreciated.

Answer 1

[Torsten]

Solve

glmval(Voltage) - (given probability) = 0

for "Voltage" using "fzero", e.g.

Best wishes

Torsten.

[Nick C]

Thanks for the feedback. Can you propose code to do this?