Example usage for weka.filters.supervised.instance SpreadSubsample setInputFormat

Introduction

In this page you can find the example usage for weka.filters.supervised.instance SpreadSubsample setInputFormat.

Prototype

@Override
public boolean setInputFormat(Instances instanceInfo) throws Exception 

Source Link

Document

Sets the format of the input instances.

Usage

From source file:mao.datamining.DataSetPair.java

private void createSampleDataSets() {
    try {/*from  www.  j a  va2 s .  co  m*/
        //reload the new data from new arff file: Main.OrangeProcessedDSHome+"/afterRemoveRows.arff"
        Instances newData = ConverterUtils.DataSource
                .read(Main.OrangeProcessedDSHome + "/afterRemoveRows2.arff");
        newData.setClassIndex(newData.numAttributes() - 1);
        //create none sample file
        //            Main.logging("== New Data After Doing Nothing, waiting for CostMatrix: ===\n" + newData.toSummaryString());
        try (BufferedWriter writer = new BufferedWriter(new OutputStreamWriter(
                new FileOutputStream(Main.OrangeProcessedDSHome + "/afterNoneSampling.arff")))) {
            writer.write(newData.toString());
        }
        //create under sample file
        //            System.out.println("Under Samplessssssssssssssssssssssssssssssssssssss");
        SpreadSubsample underSampleFilter = new weka.filters.supervised.instance.SpreadSubsample();
        underSampleFilter.setInputFormat(newData);
        String underOptionsClone[] = new String[underSampleFilterOptions.length];
        System.arraycopy(underSampleFilterOptions, 0, underOptionsClone, 0, underSampleFilterOptions.length);
        underSampleFilter.setOptions(underOptionsClone);
        Instances underNewData = Filter.useFilter(newData, underSampleFilter);

        //            Main.logging("== New Data After Under Sampling: ===\n" + underNewData.toSummaryString());
        try (BufferedWriter writer = new BufferedWriter(new OutputStreamWriter(
                new FileOutputStream(Main.OrangeProcessedDSHome + "/afterUnderSampling.arff")))) {
            writer.write(underNewData.toString());
        }
        //create over sample file
        //            System.out.println("Over Samplessssssssssssssssssssssssssssssssssssss");
        //weka.filters.supervised.instance.SMOTE -C 0 -K 5 -P 1000.0 -S 1 smoteOptions
        SMOTE smote = new weka.filters.supervised.instance.SMOTE();
        smote.setInputFormat(newData);
        String overOptionsClone[] = new String[overSampleSmoteOptions.length];
        System.arraycopy(overSampleSmoteOptions, 0, overOptionsClone, 0, overSampleSmoteOptions.length);
        smote.setOptions(overOptionsClone);
        Instances overNewData = Filter.useFilter(newData, smote);

        //            Main.logging("== New Data After Over Sampling: ===\n" + overNewData.toSummaryString());
        try (BufferedWriter writer = new BufferedWriter(new OutputStreamWriter(
                new FileOutputStream(Main.OrangeProcessedDSHome + "/afterOverSampling.arff")))) {
            writer.write(overNewData.toString());
        }

    } catch (Exception ex) {
        Logger.getLogger(DataSetPair.class.getName()).log(Level.SEVERE, null, ex);
    }
}

---