Algorithm

One Hot Encoder

Description

Encodes categorical features as a one-hot numeric array.

Implementation Details

Implemented in `encoders.R`.

transform=function(df){
  # ... setup code ...
  for(colname in names(df)){
    col <- as.character(df[[colname]])
    cats <- categories[[colname]]
    idx <- match(col, cats)
    # ... fill matrix ...
    out[cbind(row_ids, col_ids)] <- 1L
  }
  out
}

Use Cases

General purpose.