dismo MaxEnt does not run with categorical variables

Question

I am trying to build a MaxEnt model in dismo, using a data frame. I already have presence-absence data and extracting the covariates for each point, so the only data I need is this one data frame. I have no issues running the model with continuous explanatory variables, but when I try to include a categorical variable, the model fails to build.

library(dismo)

dat <- data.frame(a = c(0, 0, 1, 1, 0),
                  b = c(1, 2, 3, 4, 5),
                  d = c(7, 8, 9, 9, 8),
                  e = as.factor(c("q", "r", "s", "r", "q")))

maxent1 <- dismo::maxent(x = dat[, 2:3], p = dat$a)
#works fine

maxent2 <- dismo::maxent(x = dat[, 2:4], p = dat$a)
#does not run at all
#Error: In file(con, "r") :cannot open file 'species.lambdas': No such file or directory

maxent3 <- dismo::maxent(x = dat[, c(2:3, as.factor(4))], p = dat$a)
#runs but leaves out the factor variable

Has anyone figured out a way around this issue? I can use a RasterStack but would prefer to use a data frame as this is part of a multi-model function.

Answer 1

After looking through the manual for the actual MaxEnt program, I see that categorical variables must be represented by numbers.

This code runs just fine and includes everything as expected.

dat <- data.frame(a = c(0, 0, 1, 1, 0),
                  b = c(1, 2, 3, 4, 5),
                  d = c(7, 8, 9, 9, 8),
                  e = as.factor(c("4", "5", "6", "5", "4")))

maxent2 <- dismo::maxent(x = dat[, 2:4], p = dat$a)