Package com.jml.preprocessing

Class Encoder

java.lang.Object
com.jml.preprocessing.Encoder
public class Encoder extends Object
Contains methods to encode classes or targets to numerical values. This allows labels to be used as targets when fitting a Model.

  • Method Summary

    Modifier and Type
    Method
    Description
    static int[]
    encodeClasses​(int[] labels)
    Encodes a list of classes as values between 0 to (n-1) classes.
    static int[]
    encodeClasses​(String[] labels)
    Encodes a list of classes as values between 0 to (n-1) classes.

    Labels will be sorted lexicographically before encoding.
    static int[]
    encodeClasses​(String[][] labels)
    Encodes a list of classes as values between 0 t0 (n-1) classes.
    WARNING: This method only flattens the array and calls encodeClasses(String[]).

    Labels will be sorted alphabetically before encoding.

    Methods inherited from class java.lang.Object

    equals, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

  • Method Details

    • encodeClasses

      public static int[] encodeClasses​(int[] labels)
      Encodes a list of classes as values between 0 to (n-1) classes.

      Labels will be sorted lexicographically before encoding. This guarantees a consistent method of encoding so that the encoded labels can be decoded.
      Parameters:
      labels - Labels of a dataset.
      Returns:
      An integer array containing the encoding of each label.

    • encodeClasses

      public static int[] encodeClasses​(String[] labels)
      Encodes a list of classes as values between 0 to (n-1) classes.

      Labels will be sorted lexicographically before encoding. This guarantees a consistent method of encoding so that the encoded labels can be decoded.
      Parameters:
      labels - Labels of a dataset.
      Returns:
      An integer array containing the encoding of each label.

    • encodeClasses

      public static int[] encodeClasses​(String[][] labels)
      Encodes a list of classes as values between 0 t0 (n-1) classes.
      WARNING: This method only flattens the array and calls encodeClasses(String[]).

      Labels will be sorted alphabetically before encoding. This guarantees a consistent method of encoding so that the encoded labels can be decoded.
      Parameters:
      labels - Labels of a dataset.
      Returns:
      An integer array containing the encoding of each label.