edu.umass.cs.mallet.share.upenn.ner

Class LengthBins

java.lang.Object

edu.umass.cs.mallet.base.pipe.Pipe

edu.umass.cs.mallet.share.upenn.ner.LengthBins

All Implemented Interfaces:

Serializable

public class LengthBins
extends Pipe
implements Serializable

A feature approximating string length.

See Also:

Serialized Form

Constructor Summary

Constructors

Constructor and Description
LengthBins(String featureName, int[] binMaxes) bins contains the maximum sizes of elements in each bin.

Method Summary

Modifier and Type	Method and Description
Instance	pipe(Instance carrier) Process an Instance.

Methods inherited from class edu.umass.cs.mallet.base.pipe.Pipe

getDataAlphabet, getInstanceId, getParent, getParentRoot, getTargetAlphabet, isDataAlphabetSet, isTargetProcessing, pipe, readResolve, resolveDataAlphabet, resolveTargetAlphabet, setDataAlphabet, setParent, setTargetAlphabet, setTargetProcessing

Methods inherited from class java.lang.Object

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

Constructor Detail

LengthBins

public LengthBins(String featureName,
                  int[] binMaxes)

bins contains the maximum sizes of elements in each bin.

For example, passing in {1,3,7} would produce 4 bins, for strings of lengths 1, 2-3, 4-7, and 8+.

Method Detail

pipe

public Instance pipe(Instance carrier)

Description copied from class: Pipe

Process an Instance. This method takes an input Instance, destructively modifies it in some way, and returns it. This is the method by which all pipes are eventually run.

One can create a new concrete subclass of Pipe simply by implementing this method.

Specified by:

pipe in class Pipe

Parameters:

carrier - Instance to be processed.