mlogit package in R: with alternative specific variables and without intercept

Question

I'm new to R, and I'm trying to run a logit model with alternative specific variables and without intercept. I have checked the documentation, but it seems that when I run the model with alternative specific variables, the intercepts are always included. Could anyone kindly tell me how to run the model without intercept? Thanks. Below is what I have tried. I use the data set Fishing in the mlogit package in R.

m.fishing.1 <- mlogit(mode ~ price + catch | income, data = df.fishing, reflevel = "beach");

Len Greski · Accepted Answer · 2020-04-15T15:32:43.873

Use +0 in the model formula specification to exclude an intercept. From the help for mlogit::mFormula():

data("Fishing", package = "mlogit")
Fish <- mlogit.data(Fishing, varying = c(2:9), shape = "wide", choice =
                      "mode")    
f3 <- mFormula(mode ~ price + catch | income + 0)
mlogit(f3,Fish)

...and the output:

> mlogit(f3,Fish)

Call:
mlogit(formula = mode ~ price + catch | income + 0, data = Fish,     method = "nr")

Coefficients:
         price           catch     income:boat  income:charter     income:pier  
   -2.1597e-02      7.1208e-01      2.1014e-04      2.4498e-04      3.1272e-05  

>

mlogit package in R: with alternative specific variables and without intercept

1 Answers1