statsmodels.gam.generalized_additive_model.LogitGam.score_factor¶

LogitGam.score_factor(params)¶

Logit model derivative of the log-likelihood with respect to linpred.

Parameters:¶

paramsarray_like

The parameters of the model

Returns:¶

score_factorarray_like

The derivative of the loglikelihood for each observation evaluated at params.

Notes

$$\frac{\partial \ln L_i}{\partial \beta }=(y_i-{\lambda }_i)$$

for observations $i=1,...,n$

where the loglinear model is assumed

$$\ln {\lambda }_i=x_i\beta$$