stepwise logistic regression using statistics toolbox

The response y in "stepwiseglm" has to be one column vector?

I found that the response y in "glmfit" can be two column matrix with the first column indicating the number of successes for each observation and the second column indicating the number of trials for each observation.

My response y is in format [successes; num of trials] and I want to use "stepwiseglm". How can I do that? Thank you so much!

Boyi

Answer 1

[the cyclist]

You can specify the response variable y as a two-column vector, just as you did in glmfit. You must also then specify a binomial distribution, by adding the name-value pair like this ...

stepwiseglm(...,'Distribution','Binomial')

This is mentioned in the documentation for stepwiseglm.

[Boyi Jiang 2015]

Thank you so much! One further question: once I obtained the model, how can I use "[X,Y] = perfcurve(labels,scores,posclass) " to create the ROC curve? Since the response has two colunms: [successes; num of trials]. How to create the "labels"?

[the cyclist]

I don't know. You might want to post this as a new question, to get more visibility.