0

I'm trying to use the mlogit package in R in order to make a choice model on a panel dataset. For this, I have successfully specified the data in mlogit.data form. The y variable consists of TRUE / FALSE indicating the choice, the service variable consists of all choice options, and the id variable consists of all individuals who make the choices.

data <- mlogit.data(df, choice = "y", 
                    shape = "long",
                    alt.var = "service",
                    id.var = "id")

Considering that I get the following error:

Error in str2lang(x) : <text>:2:0: unexpected end of input
1: . ~   . | 
   ^

I'm assuming that I'm misspecifying my formula. I've tried to specify it in the following two manners, yet both result in the same error.

f1 <- mFormula(y ~ x1 + x2) # does not throw error

head(model.matrix(f1, data), 2) # throws the error

model1 <- mlogit(f1, data = data,
               rpar = c(x1 = "n", 
                        x2 = "n"),
               panel = TRUE)  # throws the error

and

f2 <- y ~ x1 + x2 # does not throw the error

head(model.matrix(f2, data), 2) # does not throw error

model2 <- mlogit(f2, data = data,
               rpar = c(x1 = "n", 
                        x2 = "n"),
               panel = TRUE)  # throws the error

The variables that I'm using in my formula are alternative-specific, meaning that for service s at period t, the variables have the same values.

Does anyone know what is causing the error? Thanks in advance!

Spectre Yen
  • 29
  • 5

1 Answers1

1

I was running into the same error. First, the mlogit.data function is deprecated, so dfidx needs to be used to set up the format of the data. Secondly, when I was setting up my data using dfidx, I was departing from a data.table object and then when attempting to fit the model, I got the error you are asking about.

The way I solved it was by preparing the dfidx data using a data.frame object.

Vfd1990
  • 11
  • 1