Compute Information matrix for model parameters.
Usage
multinom_info_mat(X, Y, probs)
Arguments
- X
This should be the \(n x p\) design matrix of covariates.
- Y
This should be the \(n x J\) data matrix of outcomes.
- probs
This should be the \(n \times J\) matrix of estimated probabilities for each sample.
Value
The Fisher information matrix for model parameters.