Coefficients of a fitted logistic regression — glm

What a GLM has that the others do not: a signed, testable number per predictor. Because the recipe centres and scales, these are on a common scale and can be read against each other — a coefficient of 0.8 moves the log-odds by 0.8 per standard deviation of its predictor.

Usage

glm_coefficients(fitted)

Arguments

fitted: a fitted workflows::workflow() whose model is a glm

Value

a data frame of variable, estimate, std_error, statistic, p_value, and the 95% interval as lower/upper

Examples

if (FALSE) { # \dontrun{
glm_coefficients(model$workflow)
} # }