5

In this post HERE they discuss how to one-hot encode a single factor variable in R. I wonder how to reverse to the problem and get a single factor from variables that one-hot encode certain properties?

striatum
  • 1,428
  • 3
  • 14
  • 31

2 Answers2

2

Here's a solution ...

First one hot encode carb

mtcars$carb <- factor(mtcars$carb)
df <- as.data.frame(model.matrix(~ carb - 1, mtcars))
head(df)

#>                   carb1 carb2 carb3 carb4 carb6 carb8
#> Mazda RX4             0     0     0     1     0     0
#> Mazda RX4 Wag         0     0     0     1     0     0
#> Datsun 710            1     0     0     0     0     0
#> Hornet 4 Drive        1     0     0     0     0     0
#> Hornet Sportabout     0     1     0     0     0     0
#> Valiant               1     0     0     0     0     0

We could of course select out the hot encode variables

library(dplyr)

df %>% 
   rowwise() %>% 
   mutate(remade = which.max(c_across(starts_with("carb")))) %>%
   ungroup %>%
   mutate(remade = factor(remade))

#> # A tibble: 32 x 7
#>    carb1 carb2 carb3 carb4 carb6 carb8 remade
#>    <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <fct> 
#>  1     0     0     0     1     0     0 4     
#>  2     0     0     0     1     0     0 4     
#>  3     1     0     0     0     0     0 1     
#>  4     1     0     0     0     0     0 1     
#>  5     0     1     0     0     0     0 2     
#>  6     1     0     0     0     0     0 1     
#>  7     0     0     0     1     0     0 4     
#>  8     0     1     0     0     0     0 2     
#>  9     0     1     0     0     0     0 2     
#> 10     0     0     0     1     0     0 4     
#> # … with 22 more rows

Here it is as a function with the option to keep or delete the one hot encoded columns a la @KM_83

cold_encode <- function(df, encoded_prefix, keep_dummies = FALSE) {
   var <- sym(encoded_prefix)
   df <- 
      df %>%
      rowwise() %>%
      mutate({{ var }} := which.max(c_across(starts_with(encoded_prefix)))) %>%
      ungroup %>%
      mutate({{ var }} := factor({{ var }})) 
   if (!keep_dummies) {
      df <- 
      df %>% select(-matches(paste0(encoded_prefix,1:9)))
   }
   return(df)
}

cold_encode(df, "carb")
#> # A tibble: 32 x 11
#>      mpg   cyl  disp    hp  drat    wt  qsec    vs    am  gear carb 
#>    <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <fct>
#>  1  21       6  160    110  3.9   2.62  16.5     0     1     4 4    
#>  2  21       6  160    110  3.9   2.88  17.0     0     1     4 4    
#>  3  22.8     4  108     93  3.85  2.32  18.6     1     1     4 1    
#>  4  21.4     6  258    110  3.08  3.22  19.4     1     0     3 1    
#>  5  18.7     8  360    175  3.15  3.44  17.0     0     0     3 2    
#>  6  18.1     6  225    105  2.76  3.46  20.2     1     0     3 1    
#>  7  14.3     8  360    245  3.21  3.57  15.8     0     0     3 4    
#>  8  24.4     4  147.    62  3.69  3.19  20       1     0     4 2    
#>  9  22.8     4  141.    95  3.92  3.15  22.9     1     0     4 2    
#> 10  19.2     6  168.   123  3.92  3.44  18.3     1     0     4 4    
#> # … with 22 more rows
Chuck P
  • 3,862
  • 3
  • 9
  • 20
1

How about this?

library(dplyr)
df <- mtcars
df$cyl <- as.factor(df$cyl) 
dummies <- data.frame(model.matrix( ~ 0 + cyl, df))  # create one-hot dummies
df <- cbind(mtcars %>% select(-cyl), dummies)

convert_one_hot_back <- function(df, prefix, keep_dummies=FALSE) {
  df_dum <- df %>% dplyr::select(starts_with(prefix))
  vars <- names(dummies) 
  df_dum[[prefix]] <- NA
  for (var in vars) {
    idx <- df_dum[[var]] == 1
    df_dum[[prefix]][idx] <- gsub(prefix,'', var)
  }
  if (!keep_dummies) {
    df_dum <- df_dum %>% select(prefix)
  }
  return(cbind(df %>% dplyr::select(-vars), df_dum ))
}

convert_one_hot_back(df, 'cyl') # convert one-hot dummies to the dummy-label var
KM_83
  • 697
  • 3
  • 9