how to get k nearest neighbors using weka kdtree

824 views Asked by At

I'm trying to get k nearest neighbors using weka KDTree implementation like this:

    ArrayList<ArrayList<Double>> ar = new ArrayList<ArrayList<Double>>();
    ArrayList<Double> d1 = new ArrayList<Double>();
    d1.add(1.1);
    d1.add(1.1);

    ArrayList<Double> d2 = new ArrayList<Double>();
    d2.add(2.2);
    d2.add(2.2);

    ArrayList<Double> d3 = new ArrayList<Double>();
    d3.add(3.3);
    d3.add(3.3);

    ar.add(d1);
    ar.add(d2);
    ar.add(d3);

    Attribute a1 = new Attribute("attr1", 0);
    Attribute a2 = new Attribute("attr2", 0);
    FastVector attrs = new FastVector();
    attrs.addElement(a1);
    attrs.addElement(a2);

    Instances ds = new Instances("ds", attrs, 10);
    for (ArrayList<Double> d : ar) {
        Instance i = new Instance(2);
        i.setValue(a1, d.get(0));
        i.setValue(a2, d.get(1));
        ds.add(i);
    }
    Instance target = new Instance(2);
    target.setValue(a1, 7);
    target.setValue(a2, 7);
    KDTree knn = new KDTree(ds);

    Instances targetDs = new Instances("target", attrs, 1);
    targetDs.add(target);

    Instances nearestInstances = knn.kNearestNeighbours(targetDs.firstInstance(), 2);
    for (int i = 0; i < nearestInstances.numInstances(); i++) {

        System.out.println(nearestInstances.instance(i).value(a1) + ", "
                + nearestInstances.instance(i).value(a2));
    }

But it throws a NullPointerException in kNearestNeighbours call:

Exception in thread "main" java.lang.NullPointerException at weka.core.neighboursearch.KDTree.findNearestNeighbours(KDTree.java:308) at weka.core.neighboursearch.KDTree.kNearestNeighbours(KDTree.java:390) at blah.App.main(App.java:60)

I couldn't find any hint in docs and the exception message is of no use. Any idea what might be the problem here?

1

There are 1 answers

3
Mehraban On BEST ANSWER

Well, using constructor without parameter and setting the param in next step solved the issue here. I mean I changed

    KDTree knn = new KDTree(ds);

to

    KDTree knn = new KDTree();
    knn.setInstances(ds);

and it works. I don't know what to tell, just congrats weka!