n-1 Dummies

[benjaminschlegel]

Thanks a lot for the great package, it's very usefull.

Is there a possibility you could add an option that removes the first dummy of every variable that only n-1 Dummies remain? It would be very helpfull for models which need numeric data, but only n-1 Dummies because of perfect multicolinearity.

The code for it:

• add variable "remove.first.dummy = FALSE" as parameter to the function (or however you want to call it)

• add the code below:

  for (col_name in char_cols) { unique_vals <- unique(dataset[, get(col_name)])

  ----- next lines new ---

  if(remove.first.dummy){ unique_vals = unique_vals[-1] }

  -------------------------

  dataset[, (paste0(colname, "", unique_vals)) := 0] for (unique_values in unique_vals) { dataset[get(col_name) == unique_values, (paste0(colname, "", unique_values)) := 1] } }

[jacobkap]

Thanks for your contribution! I'll update the package soon. Would you like to be added as a contributor to the package?

[benjaminschlegel]

If it's not much work for you to add me, that would be nice.

[jacobkap]

Can you send me your email? I can add that to the documentation.

[benjaminschlegel]

kontakt@benjaminschlegl.ch