A simple function to generate both data file and description file for the regression tree software GUIDE

prepare_for_guide.R

# We want to create clean data file and description file for GUIDE
prepare_for_guide = function(df, y_name, output_name, desc_name){
    # Find the index of response variable y
    y_index = which(colnames(df)==y_name)
    
    # Data file
    write.table(df, output_name, row.names=FALSE, col.names=FALSE, quote=FALSE)
    
    # Description file
    desc_file = file(desc_name)
    lines = c(output_name, "NA", 1)
    names = colnames(df)
    
    for (i in 1:length(names)){
        # Check if discrete variable or response variable
        if (i == y_index){
            column_type = 'd'
        } else if (is.factor(df[,i])) {
            column_type = 'c'
        } else {
            column_type = 'n'
        }
        lines[length(lines)+1] = sprintf("%d %s %s", i, names[i], column_type)
    }

    writeLines(lines, desc_file)
    close(desc_file)
}