scikits.statsmodels.discretemod.MNLogit.fit

MNLogit.fit(start_params=None, maxiter=35, method='newton', tol=1e-08)

Fits the multinomial logit model.

Parameters :

start_params : array-like, optional

The default is a 0 vector.

maxiter : int, optional

Maximum number of iterations. The default is 35.

method : str, optional

method can be ‘newton’, ‘ncg’, ‘bfgs’. The default is ‘newton’.

tol : float, optional

The convergence tolerance for the solver. The default is 1e-08.

Notes

The reference category is always the first column of wendog for now.

Previous topic

scikits.statsmodels.discretemod.MNLogit.cdf

Next topic

scikits.statsmodels.discretemod.MNLogit.hessian

This Page