Example usage for weka.core Instance setValue

Introduction

In this page you can find the example usage for weka.core Instance setValue.

Prototype

public void setValue(Attribute att, String value);

Source Link

Document

Sets a value of an nominal or string attribute to the given value.

Usage

From source file:processes.ClusterProcess.java

private Instance toInstance(User user, Instances dataSet) {
    Instance tempInstance = new Instance(userPoints.numAttributes());
    tempInstance.setDataset(userPoints);
    String userDataString = "0" + user.getTasteString(LastFMDataHandler.getInitialTagCount());
    String[] dataArray = userDataString.split(",");
    for (int index = 0; index < dataArray.length; index++) {
        tempInstance.setValue(index, Integer.parseInt(dataArray[index]));
    }// w  ww .  j  a v a  2 s .  co  m
    return tempInstance;
}

---

From source file:put.semantic.fcanew.ml.WekaClassifier.java

public static Instance convert(Instance input, Instances src, Instances dst) {
    Instance result = new Instance(dst.numAttributes());
    result.setDataset(dst);// w  w  w.ja v  a 2  s  .  c  o  m
    for (int i = 0; i < dst.numAttributes(); ++i) {
        Attribute srcAttr = src.attribute(dst.attribute(i).name());
        if (srcAttr.isNumeric()) {
            double val = input.value(srcAttr);
            result.setValue(i, val);
        } else {
            String val = input.stringValue(srcAttr);
            result.setValue(i, val);
        }
    }
    return result;
}

---

From source file:qa.experiment.ProcessFeatureVector.java

public String trainAndPredict(String[] processNames, String question) throws Exception {
    FastVector fvWekaAttribute = generateWEKAFeatureVector(processNames);
    Instances trainingSet = new Instances("Rel", fvWekaAttribute, bowFeature.size() + 1);
    trainingSet.setClassIndex(bowFeature.size());

    int cnt = 0;/*from   ww w.  j a v a2  s  .c o  m*/
    for (int i = 0; i < arrProcessFeature.size(); i++) {
        String[] names = arrProcessFeature.get(i).getProcessName().split("\\|");
        int sim = isNameFuzzyMatch(processNames, names);
        if (sim != -1) {
            // System.out.println("match " + arrProcessFeature.get(i).getProcessName());
            ArrayList<String> featureVector = arrProcessFeature.get(i).getFeatureVectors();
            for (int j = 0; j < featureVector.size(); j++) {
                Instance trainInstance = new Instance(bowFeature.size() + 1);
                String[] attrValues = featureVector.get(j).split("\t");
                // System.out.println(trainInstance.numAttributes());
                // System.out.println(fvWekaAttribute.size());
                for (int k = 0; k < bowFeature.size(); k++) {
                    trainInstance.setValue((Attribute) fvWekaAttribute.elementAt(k),
                            Integer.parseInt(attrValues[k]));
                }
                trainInstance.setValue((Attribute) fvWekaAttribute.elementAt(bowFeature.size()),
                        processNames[sim]);
                trainingSet.add(trainInstance);

                //System.out.println(cnt);
                cnt++;
            }
        }
    }

    Classifier cl = new NaiveBayes();
    cl.buildClassifier(trainingSet);
    Instance inst = new Instance(bowFeature.size() + 1);
    //String[] tokenArr = tokens.toArray(new String[tokens.size()]);
    for (int j = 0; j < bowFeature.size(); j++) {
        List<String> tokens = slem.tokenize(question);
        String[] tokArr = tokens.toArray(new String[tokens.size()]);
        int freq = getFrequency(bowFeature.get(j), tokArr);
        inst.setValue((Attribute) fvWekaAttribute.elementAt(j), freq);
    }

    inst.setDataset(trainingSet);
    int idxMax = ArrUtil.getIdxMax(cl.distributionForInstance(inst));
    return processNames[idxMax];
}

---

From source file:qa.qcri.nadeef.core.utils.classification.ClassifierBase.java

License:Open Source License

/**
 * Update the existing classifier with new instance. For online models, it directly updates. For offline learning models, it re-generates the model with updated training set
 *
 * @param instance/*from  w w w.j a v  a 2  s  .c o  m*/
 */
public void updateClassifier(TrainingInstance instance) throws NadeefClassifierException {
    // transform training instance into real instance
    Instance wekaInstance = new Instance(numberOfAttributes);
    wekaInstance.setDataset(instances);
    // add values from old tuple
    for (Cell cell : instance.getDirtyTuple().getCells()) {
        if (isPermitted(cell.getColumn())) {
            if (isPermitted(cell.getColumn())) {
                if (cell.getValue() instanceof String) {
                    wekaInstance.setValue(attributeIndex.get(cell.getColumn()), cell.getValue().toString());
                } else {
                    double doubleValue = Double.parseDouble(cell.getValue().toString());
                    wekaInstance.setValue(attributeIndex.get(cell.getColumn()), doubleValue);
                }
            }
        }
    }

    // add new value, check its type from dirty value
    if (instance.getDirtyTuple().getCell(instance.getAttribute()).getValue() instanceof String) {
        wekaInstance.setValue(numberOfAttributes - 3, instance.getUpdatedValue());
    } else {
        double doubleValue = Double.parseDouble(instance.getUpdatedValue());
    }
    // add similarity
    wekaInstance.setValue(numberOfAttributes - 2, instance.getSimilarityScore());
    // add class label
    wekaInstance.setValue(numberOfAttributes - 1, instance.getLabel().toString());

    updateClassifier(wekaInstance);
}

---

From source file:qa.qcri.nadeef.core.utils.classification.ClassifierBase.java

License:Open Source License

/**
 * Get Prediction for a given instance based on current model
 *
 * @param instance//  www .ja v  a 2  s  . co m
 */
public ClassificationResult getPrediction(TrainingInstance instance) throws NadeefClassifierException {
    // transform training instance into real instance
    Instance wekaInstance = new Instance(numberOfAttributes);
    wekaInstance.setDataset(instances);
    // add values from old tuple
    for (Cell cell : instance.getDirtyTuple().getCells()) {
        if (isPermitted(cell.getColumn())) {
            if (cell.getValue() instanceof String) {
                wekaInstance.setValue(attributeIndex.get(cell.getColumn()), cell.getValue().toString());
            } else {
                double doubleValue = Double.parseDouble(cell.getValue().toString());
                wekaInstance.setValue(attributeIndex.get(cell.getColumn()), doubleValue);
            }
        }
    }

    // add new value, check its type from the dirty value
    if (instance.getDirtyTuple().getCell(instance.getAttribute()).getValue() instanceof String) {
        wekaInstance.setValue(numberOfAttributes - 3, instance.getUpdatedValue());
    } else {
        double doubleValue = Double.parseDouble(instance.getUpdatedValue());
    }
    // add similarity
    wekaInstance.setValue(numberOfAttributes - 2, instance.getSimilarityScore());

    double[] result = getPrediction(wekaInstance);
    // now convert this result into readable form
    ClassificationResult classificationResult = new ClassificationResult(result,
            wekaInstance.attribute(this.numberOfAttributes - 1));
    return classificationResult;
}

---

From source file:qoala.arff.java

public void valueSession(setData f) {
    vals = new double[cl.Datapoints.size() / 5][6];
    //set up Attributes//
    attributes = new FastVector();

    attributes.addElement(new Attribute("att1 SessionID"));
    attributes.addElement(new Attribute("att2 NumberofClicks"));
    attributes.addElement(new Attribute("att3 NumberofQueriesInSession"));
    attributes.addElement(new Attribute("att4 NumberofQueryTermsInSession"));
    attributes.addElement(new Attribute("att5 NumberofAverageQueryTermsInSession"));
    attributes.addElement(new Attribute("att6 SessionDuration"));
    //Creates instances object//
    dataSet = new Instances("My weka", attributes, 0);

    // fills the instances with data//
    for (int k = 0; k < cl.Datapoints.get(cl.Datapoints.size() / 5).getSession(); k++) {

        vals[k][0] = (double) f.GetSessionRetrieval(k);

        vals[k][1] = (double) f.GetClick(k);

        vals[k][2] = (double) f.GetSessionNumber(k);

        vals[k][3] = (double) f.GetTerms(k);

        vals[k][4] = (double) f.GetAverageTerms(k);

        vals[k][5] = (double) f.sessionDurationRetrieval(k);

        Instance inst = new DenseInstance(6);
        inst.setValue(0, vals[k][0]);
        inst.setValue(1, vals[k][1]);/*from   ww w .  j a  v  a 2  s .  c om*/
        inst.setValue(2, vals[k][2]);
        inst.setValue(3, vals[k][3]);
        inst.setValue(4, vals[k][4]);
        inst.setValue(5, vals[k][5]);

        dataSet.add(inst);

    }

}

---

From source file:reactivetechnologies.sentigrade.dto.RequestData.java

License:Apache License

/**
 * /*from ww  w  . jav  a 2 s. c om*/
 * @param struct
 * @param t
 * @return
 */
protected Instance buildInstance(Instances struct, Tuple t) {
    Instance i = new DenseInstance(2);
    i.setDataset(struct);
    i.setValue(struct.attribute(ClassificationModelEngine.CLASSIFIER_ATTRIB_TEXT_IDX), t.getText());
    i.setValue(struct.attribute(ClassificationModelEngine.CLASSIFIER_ATTRIB_CLASS_IDX), t.getTextClass());
    return i;
}

---

From source file:Reader.Classifyer.java

/**
 * Converts a list of images into instances
 * @param images//from  ww w .  ja  v a2s  . c  om
 * @return 
 */
private static Instances ImageListToInstances(List<BufferedImage> images) {
    int letterCount = 0;
    FastVector letters = new FastVector();
    for (String thisLetter : letterVal) {
        letters.addElement(thisLetter);
    }

    FastVector fvNominalVal = new FastVector(ATT_COUNT + 1);
    for (Integer x = 0; x < ATT_COUNT + 1; x++) {
        fvNominalVal.addElement(new Attribute(x.toString()));
    }

    Instances trainingData = new Instances("letters", fvNominalVal, 6500);
    trainingData.setClassIndex(ATT_COUNT);

    for (BufferedImage image : images) {
        Instance temp = imageToInstance(image);
        try {
            temp.setValue(ATT_COUNT, letterCount++ / 24);
        } catch (Exception ex) {
            System.out.println(ex + " Thrown in Classifying");
        }

        trainingData.add(temp);
    }

    return trainingData;
}

---

From source file:Reader.Classifyer.java

/**
 * Converts the image into an instance by taking the value of each pixel, the
 * number of black pixels in each row/col of the image, and the number of strokes
 * on each row/col.//ww w.j a v a2  s .c  o m
 * @param image - The image to convert
 * @return The resulting Instance
 */
private static Instance imageToInstance(BufferedImage image) {
    BufferedImage newImage = resize(image);
    int height = newImage.getHeight();
    int width = newImage.getWidth();
    List<Integer> gaps = new ArrayList<>();
    WritableRaster raster = newImage.getRaster();

    int[] pixels = new int[1537];
    raster.getPixels(0, 0, width, height, pixels);

    Instance imageInstance = new Instance(ATT_COUNT + 1);

    for (int i = 0; i < 513; i++) {
        imageInstance.setValue(i, pixels[i] / 250);
    }

    int sum;
    for (int i = 0; i < height; i++) {
        sum = 0;
        for (int j = 0; j < width; j++) {
            sum += pixels[(i * width) + j];
        }

        imageInstance.setValue(i + 512, sum / 255);
    }

    for (int i = 0; i < width; i++) {
        sum = 0;
        for (int j = 0; j < height; j++) {
            sum += pixels[(j * width) + i];
        }

        imageInstance.setValue(i + height + 512, sum / 255);
    }

    int offset = 512 + height + width;

    // Extra weight added to the number of strokes because it is less likely
    // to be effected by noise in the image
    // Strokes across the x
    int[] rowPixels = new int[width * 3];
    for (int i = 0; i < height; i++) {
        raster.getPixels(0, i, width, 1, rowPixels);
        imageInstance.setValue(i + offset, getStrokes(rowPixels) * 20);
    }

    offset += height;

    // Strokes accross the y
    int[] colPixels = new int[height * 3];
    for (int i = 0; i < width; i++) {
        raster.getPixels(i, 0, 1, height, colPixels);
        imageInstance.setValue(i + offset, getStrokes(colPixels) * 20);
    }

    return imageInstance;
}

---

From source file:ru.ksu.niimm.cll.mocassin.crawl.analyzer.relation.impl.WekaBasedNavigationalRelationClassifier.java

License:Open Source License

/**
 * {@inheritDoc}//from   ww  w  .  ja va  2  s.c o m
 */
@Override
public Prediction predict(Reference reference, Graph<StructuralElement, Reference> graph) {
    StructuralElement from = graph.getSource(reference);
    MocassinOntologyClasses fromType = from.getPredictedClass();
    StructuralElement to = graph.getDest(reference);
    MocassinOntologyClasses toType = to.getPredictedClass();
    long documentSize = reference.getDocument().getSize();
    float normalizedStartDistance = ((float) from.getGateStartOffset() - to.getGateStartOffset())
            / documentSize;

    float normalizedEndDistance = ((float) from.getGateEndOffset() - to.getGateEndOffset()) / documentSize;
    Instance instance = new Instance(trainingSetHeader.numAttributes());
    instance.setDataset(trainingSetHeader);
    instance.setValue(0, fromType.toString());
    instance.setValue(1, toType.toString());
    instance.setValue(2, normalizedStartDistance);
    instance.setValue(3, normalizedEndDistance);
    for (int i = 4; i < trainingSetHeader.numAttributes() - 1; i++) {
        String attrName = trainingSetHeader.attribute(i).name();
        String word = attrName.substring(attrName.indexOf(ATTRIBUTE_NAME_DELIMITER) + 1);
        instance.setValue(i, reference.getSentenceTokens().contains(word) ? 1 : 0);
    }
    try {
        double[] distribution = classifier.distributionForInstance(instance);
        Prediction prediction;
        if (distribution[0] > distribution[1]) {
            prediction = new Prediction(MocassinOntologyRelations.REFERS_TO, distribution[0]);
        } else {
            prediction = new Prediction(MocassinOntologyRelations.DEPENDS_ON, distribution[1]);
        }
        return prediction;
    } catch (Exception e) {
        logger.error("Couldn't classify a reference with id='{}' in a document='{}'; null will be returned",
                format("%d/%s", reference.getId(), reference.getDocument().getUri()), e);
        return null;
    }
}

---

• «
• 1
• 2
• 3
• 4
• 5
• 6
• 7
• 8
• 9
• 10
• 11
• 12
• 13
• 14
• 15
• 16
• 17
• 18
• 19
• 20
• 21
• 22
• 23
• 24
• 25
• 26
• 27
• 28
• 29
• 30
• 31
• 32
• 33
• 34
• 35
• »