Function to generate dummy variables from the possible sets of values

create_value_dummies.R

#########################################################
# Define function: create_value_dummies
# Create dummy variables indicating if a given
# profile takes on a specific value from its 
# possible values. Create a dummy for the range
# of the variable, independent of whether the actual
# random sampling does (not) draw that value and assign
# it to a profile.
#########################################################

create_value_dummies <-
  function(attr_df,
           attr_levels) {
    # Convert attribute levels to a list for
    # easing referencing if not already list
    if (!is.list(attr_levels)) {
      attr_levels <- list(attr_levels)
    }
    
    # Initialize list to store output data frame
    attr_cols <- list()
    
    # Generate dummy variables
    for (i in 1:ncol(attr_df)) {
      # List to store current attribute's
      # output columns
      temp <- list()
      
      # Levels of current attribute
      this_attr_levels <- attr_levels[[i]]
      
      # Generate dummies
      for (j in 1:length(this_attr_levels)) {
        temp[[j]] <- ifelse(attr_df[, i] == attr_levels[[i]][j], 1, 0)
      }
      
      # Convert each attribute's vectors from list to data frame
      temp <- dplyr::bind_cols(temp)
      
      # Generate column names for each
      
      # Add variable names to columns
      colnames(temp) <-
        paste0(colnames(attr_df)[i], "_", attr_levels[[i]])
      
      # Save all dummies for each attribute into an element
      # of the attr_cols list
      attr_cols[[i]] <- temp
    }
    
    # Collapse each attribute's columns into a single
    # data frame, return
    dummy_df <- dplyr::bind_cols(attr_cols)
    
    # Joint the columns to the original data
    dplyr::bind_cols(attr_df, dummy_df)
  }

###################################
# Example
###################################

# Generate example data: 2 attributes,
# different number of levels per 
# attribute
#df <- data.frame(foo = rep(1:4,10), bar = rep(c("a","b"),20))
#dflevels <- list(c(1:4), c("a","b"))

# Produce dummies
#create_dummies(attr_df = df, attr_levels = dflevels)