rmcfs (version 1.2.5)

fix.data: Fixes input data values, column names and attributes types

Description

Fixes any input data to prepare them to export to ARFF/ADX formats. If after exporting data to ARFF/ADX formats there are some problems in running Java MCFS or WEKA, try to use this function before. This function fixes data values (e.g. space " " is replaced by "_") and data types (e.g. all Date columns converted to character in R).

Usage

fix.data(x, 
          type = c("all", "names", "values", "types"), 
          source_chars = c(" ", ",", "/", "|", "#", "-", "(", ")"), 
          destination_char = "_", 
          numeric_class = c("difftime"), 
          nominal_class = c("factor", "logical", "Date", "POSIXct", "POSIXt"))

Arguments

x

input data frame to be fixed.

type

• all - fixes: column names, data values, data types.

• names - fixes only column names. All characters determined by source_chars parameter are replaced by destination_char (e.g. space " " is replaced by "_").

• values - fixes only data values. All characters determined by source_chars parameter are replaced by destination_char (e.g. space " " is replaced by "_").

• types - fixes only data types (e.g. all possible nominal columns as (Date or logical) converted to character).

source_chars

characters that will be replaced in column names and data values.

destination_char

character that will be inserted in column names and data values.

numeric_class

vector of class labels to be casted as.numeric.

nominal_class

vector of class labels to be casted as.character.

Value

data.frame with fixed values and types (depends on type parameter).

Examples

  ### Set up java parameter and load rmcfs package
  options(java.parameters = "-Xmx4g")
  library(rmcfs)
  
  # Load alizadeh dataset.
  data(alizadeh)
  alizadeh <- alizadeh[1:500]
  
  # Fix data types and data values - remove "," " " "/" from values and fix data types
  # This function may help if mcfs has any problems with input data
  alizadeh.fixed <- fix.data(alizadeh)