net.librec.data.splitter

Class RatioDataSplitter

java.lang.Object

net.librec.conf.Configured

net.librec.data.splitter.AbstractDataSplitter

net.librec.data.splitter.RatioDataSplitter

All Implemented Interfaces:

Configurable, DataSplitter

public class RatioDataSplitter
extends AbstractDataSplitter

Ratio Data Splitter.
Split dataset into train set, test set, valid set by ratio.

Author:

WangYuFeng and Liuxz

Nested Class Summary

Nested classes/interfaces inherited from interface net.librec.data.DataSplitter

DataSplitter.SplitterType

Field Summary

Fields inherited from class net.librec.data.splitter.AbstractDataSplitter

dataConvertor, LOG, testMatrix, trainMatrix, validationMatrix

Fields inherited from class net.librec.conf.Configured

conf, CONF_DATA_COLUMN_FORMAT, CONF_DATA_INPUT_PATH, CONF_DFS_DATA_DIR

Constructor Summary

Constructors

Constructor and Description
RatioDataSplitter() Empty constructor.
RatioDataSplitter(DataConvertor dataConvertor, Configuration conf) Initializes a newly created RatioDataSplitter object with convertor and configuration.

Method Summary

Methods

Modifier and Type	Method and Description
void	getFixedRatioByUser(double ratio) Split ratings into two parts: the training set consisting of user-item ratings where a fixed number of ratings corresponding to the given ratio are preserved for each user as training data with the rest as test.
void	getRatio(double trainRatio, double validationRatio) Split the rating into : (train-ratio) training, (validation-ratio) validation, and test three subsets.
void	getRatioByItem(double ratio) Split ratings into two parts: the training set consisting of user-item ratings where ratio percentage of ratings are preserved for each item, and the rest are used as the testing data.
void	getRatioByItemDate(double ratio) Split the ratings of each item (by date) into two parts: (ratio) training, (1-ratio) test subsets.
void	getRatioByRating(double ratio) Split ratings into two parts: (ratio) training, (1-ratio) test subsets.
void	getRatioByRatingDate(double ratio) Split the ratings (by date) into two parts: (ratio) training, (1-ratio) test subsets.
void	getRatioByUser(double ratio) Split ratings into two parts: the training set consisting of user-item ratings where ratio percentage of ratings are preserved for each user, and the rest are used as the testing data.
void	getRatioByUserDate(double ratio) Split the ratings of each user (by date) into two parts: (ratio) training, (1-ratio) test subsets
void	splitData() Split the dataset according to the configuration file.

Methods inherited from class net.librec.data.splitter.AbstractDataSplitter

getTestData, getTrainData, getValidData, setDataConvertor

Methods inherited from class net.librec.conf.Configured

getConf, setConf

Methods inherited from class java.lang.Object

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

Constructor Detail

RatioDataSplitter

public RatioDataSplitter()

Empty constructor.

RatioDataSplitter

public RatioDataSplitter(DataConvertor dataConvertor,
                 Configuration conf)

Initializes a newly created RatioDataSplitter object with convertor and configuration.

Parameters:

dataConvertor - the convertor for the splitter.

conf - the configuration for the splitter.

Method Detail

splitData

public void splitData()
               throws LibrecException

Split the dataset according to the configuration file.

Throws:

LibrecException - if error occurs

getRatioByRating

public void getRatioByRating(double ratio)

Split ratings into two parts: (ratio) training, (1-ratio) test subsets.

Parameters:

ratio - the ratio of training data over all the ratings

getRatioByRatingDate

public void getRatioByRatingDate(double ratio)

Split the ratings (by date) into two parts: (ratio) training, (1-ratio) test subsets.

Parameters:

ratio - the ratio of training data

getRatioByUser

public void getRatioByUser(double ratio)

Split ratings into two parts: the training set consisting of user-item ratings where ratio percentage of ratings are preserved for each user, and the rest are used as the testing data.

Parameters:

ratio - the ratio of training data

getFixedRatioByUser

public void getFixedRatioByUser(double ratio)

Split ratings into two parts: the training set consisting of user-item ratings where a fixed number of ratings corresponding to the given ratio are preserved for each user as training data with the rest as test.

Parameters:

ratio - the ratio of training data

getRatioByUserDate

public void getRatioByUserDate(double ratio)

Split the ratings of each user (by date) into two parts: (ratio) training, (1-ratio) test subsets

Parameters:

ratio - the ratio of train data

getRatioByItem

public void getRatioByItem(double ratio)

Split ratings into two parts: the training set consisting of user-item ratings where ratio percentage of ratings are preserved for each item, and the rest are used as the testing data.

Parameters:

ratio - the ratio of training data

getRatioByItemDate

public void getRatioByItemDate(double ratio)

Split the ratings of each item (by date) into two parts: (ratio) training, (1-ratio) test subsets.

Parameters:

ratio - the ratio of training data

getRatio

public void getRatio(double trainRatio,
            double validationRatio)

Split the rating into : (train-ratio) training, (validation-ratio) validation, and test three subsets.

Parameters:

trainRatio - training ratio

validationRatio - validation ratio