org.knowceans.corpus
Interface ISplitCorpus

All Known Implementing Classes:

LabelNumCorpus, NumCorpus

public interface ISplitCorpus

ISplitCorpus allows a corpus to resize and split a cross validation data set.

Author:

gregor

Method Summary

int[][]	getOrigDocIds() get the original ids of documents
ICorpus	getTestCorpus() called after split()
ICorpus	getTrainCorpus() called after split()
void	split(int order, int split, java.util.Random rand) splits two child corpora of size 1/nsplit off the original corpus, which itself is left unchanged (except storing the splits).

Method Detail

split

void split(int order,
           int split,
           java.util.Random rand)

splits two child corpora of size 1/nsplit off the original corpus, which itself is left unchanged (except storing the splits). The corpora can be retrieved using getTrainCorpus and getTestCorpus after using this function.

Parameters:

order - number of partitions

split - 0-based split of corpus returned

rand - random source (null for reusing existing splits)

getTrainCorpus

ICorpus getTrainCorpus()

called after split()

Returns:

the training corpus according to the last splitting operation

getTestCorpus

ICorpus getTestCorpus()

called after split()

Returns:

the test corpus according to the last splitting operation

getOrigDocIds

int[][] getOrigDocIds()

get the original ids of documents

Returns:

[training documents, test documents]