Converting factors to binary in R

Question

I am trying to convert a factor variable into binary / boolean (0 or 1).

Sample data:

df  <-data.frame(a = c(1,2,3), b = c(1,1,2), c = c(\"Rose\",         


        

Answer 1

You can do this with reshaping:

library(dplyr)
library(tidyr)

df %>%
  mutate(value = 1,
         c = paste0("Is", c)) %>%
  spread(c, value, fill = 0)

Answer 2

In base R, you could use sapply() on the levels, using == to check for presence and as.integer() to coerce it to binary.

cbind(df[1:2], sapply(levels(df$c), function(x) as.integer(x == df$c)), df[4])
#   a b Pink Red Rose d
# 1 1 1    0   0    1 2
# 2 2 1    1   0    0 3
# 3 3 2    0   1    0 4

But since you have a million rows, you may want to go with data.table.

library(data.table)
setDT(df)[, c(levels(df$c), "c") := 
    c(lapply(levels(c), function(x) as.integer(x == c)), .(NULL))]

which gives

df
#    a b d Pink Red Rose
# 1: 1 1 2    0   0    1
# 2: 2 1 3    1   0    0
# 3: 3 2 4    0   1    0

And you can reset the column order if you need to with setcolorder(df, c(1, 2, 4:6, 3)).

Answer 3

dummy <- function(df) {  
  NUM <- function(dataframe)dataframe[,sapply(dataframe,is.numeric)]
  FAC <- function(dataframe)dataframe[,sapply(dataframe,is.factor)]

  require(ade4)
  if (is.null(ncol(NUM(df)))) {
      DF <- data.frame(NUM(df), acm.disjonctif(FAC(df)))
      names(DF)[1] <- colnames(df)[which(sapply(df, is.numeric))]
  } else {
      DF <- data.frame(NUM(df), acm.disjonctif(FAC(df)))
  }
  return(DF)
} 

Answer 4

Using dplyr and putting it on pipe. @bramtayl's answer was cleaner but I couldn't find a way to use custom variable name. This is less clean but more DRY

expand_factor <- function(df,variable){
    variable = as.name(variable)
    paste0('~ ',variable,' -1',collapse = '') %>% 
        as.formula ->formulae

    current.na.action <- options('na.action')
    options(na.action='na.pass')
    expanded<-model.matrix(data=df,object = formulae)
    options(na.action=current.na.action)

    colnames(expanded) <-gsub(replacement = 'is_',x = colnames(expanded),pattern=variable) 

    expanded %>% 
        tbl_df %>% 
        mutate_each(funs(as.integer)) ->expanded

    return(bind_cols(df,expanded))
}

library(dplyr)
df  <-data_frame(x = iris$Species,y = iris$Petal.Width)
df <- rbind(data_frame(x=NA,y = NA),df)

df %>% 
    expand_factor('x')

> df %>% 
+   expand_factor('x')
# A tibble: 151 <U+00D7> 5
        x     y is_setosa is_versicolor is_virginica
    <chr> <dbl>     <int>         <int>        <int>
1    <NA>    NA        NA            NA           NA
2  setosa   0.2         1             0            0
3  setosa   0.2         1             0            0
4  setosa   0.2         1             0            0
5  setosa   0.2         1             0            0
6  setosa   0.2         1             0            0
7  setosa   0.4         1             0            0
8  setosa   0.3         1             0            0
9  setosa   0.2         1             0            0
10 setosa   0.2         1             0            0
# ... with 141 more rows