Figure out how mnl auto_generates dummy variables for categorical variables in the model expression. The current model.fit doesn't omit a referent category from the model. First, we need to know if the model estimation result is actually correct in this case or not. Second, if model estimation is correct how should we interpret the coefficients for different categories?
Figure out how mnl auto_generates dummy variables for categorical variables in the model expression. The current model.fit doesn't omit a referent category from the model. First, we need to know if the model estimation result is actually correct in this case or not. Second, if model estimation is correct how should we interpret the coefficients for different categories?