Merge branch 'SPM-13'
This commit is contained in:
commit
b0d886ab15
|
@ -0,0 +1,101 @@
|
|||
package de.hsel.spm.baudas.analysis;
|
||||
|
||||
import weka.clusterers.SimpleKMeans;
|
||||
import weka.core.Instances;
|
||||
import weka.filters.Filter;
|
||||
import weka.filters.unsupervised.attribute.Remove;
|
||||
|
||||
import java.io.File;
|
||||
import java.util.HashMap;
|
||||
import java.util.Map;
|
||||
|
||||
/**
|
||||
* Cluster Analysis.
|
||||
*
|
||||
* @author Johannes Theiner
|
||||
* @version 0.1
|
||||
*/
|
||||
public class Cluster implements Analysis<Map<Integer, Map<String, String>>> {
|
||||
|
||||
private Instances instances;
|
||||
private Map<Integer, Map<String, String>> result;
|
||||
|
||||
public Cluster(File file) {
|
||||
instances = load(file);
|
||||
}
|
||||
|
||||
/**
|
||||
* get result of cluster analysis.
|
||||
*
|
||||
* @return Result
|
||||
*/
|
||||
@Override
|
||||
public Map<Integer, Map<String, String>> getResult() {
|
||||
if(result == null) {
|
||||
result = new HashMap<>();
|
||||
//TODO: anpassen wenn #SPM-17 gemerged ist.
|
||||
int[] keepIndexes = new int[]{0, 1, 3, 5, 6, 7};
|
||||
Remove remove = new Remove();
|
||||
|
||||
try {
|
||||
remove.setAttributeIndicesArray(keepIndexes);
|
||||
remove.setInvertSelection(true);
|
||||
remove.setInputFormat(instances);
|
||||
instances = Filter.useFilter(instances, remove);
|
||||
} catch (Exception e) {
|
||||
e.printStackTrace();
|
||||
}
|
||||
|
||||
//creating a single cluster to get average, weka has no way to get that from a bigger one.
|
||||
SimpleKMeans fullMeans = new SimpleKMeans();
|
||||
Instances fullCentroids = null;
|
||||
try {
|
||||
fullMeans.setNumClusters(1);
|
||||
fullMeans.setPreserveInstancesOrder(true);
|
||||
fullMeans.buildClusterer(instances);
|
||||
fullCentroids = fullMeans.getClusterCentroids();
|
||||
|
||||
} catch (Exception ex) {
|
||||
ex.printStackTrace();
|
||||
}
|
||||
assert fullCentroids != null;
|
||||
|
||||
//creating real cluster
|
||||
SimpleKMeans kMeans = new SimpleKMeans();
|
||||
try {
|
||||
kMeans.setNumClusters(5);
|
||||
kMeans.setPreserveInstancesOrder(true);
|
||||
kMeans.buildClusterer(instances);
|
||||
|
||||
int count = 0;
|
||||
count = putIntoMap(fullCentroids, count);
|
||||
|
||||
Instances centroids = kMeans.getClusterCentroids();
|
||||
putIntoMap(centroids, count);
|
||||
|
||||
} catch (Exception ex) {
|
||||
ex.printStackTrace();
|
||||
}
|
||||
}
|
||||
return result;
|
||||
}
|
||||
|
||||
/**
|
||||
* puts data into map.
|
||||
*
|
||||
* @param centroids cluster analysis result
|
||||
* @param count current insert count
|
||||
* @return count increment
|
||||
*/
|
||||
private int putIntoMap(Instances centroids, int count) {
|
||||
for (int i = 0; i < centroids.numInstances(); i++) {
|
||||
Map<String, String> map = new HashMap<>();
|
||||
for (int j = 0; j < centroids.numAttributes(); j++) {
|
||||
map.put(centroids.attribute(j).name(), centroids.instance(i).stringValue(j));
|
||||
}
|
||||
result.put(count, map);
|
||||
count++;
|
||||
}
|
||||
return count;
|
||||
}
|
||||
}
|
|
@ -0,0 +1,64 @@
|
|||
package de.hsel.spm.baudas.analysis;
|
||||
|
||||
import org.junit.jupiter.api.Test;
|
||||
|
||||
import java.io.File;
|
||||
import java.net.URL;
|
||||
import java.time.Duration;
|
||||
import java.util.Map;
|
||||
import java.util.concurrent.atomic.AtomicReference;
|
||||
|
||||
import static org.junit.jupiter.api.Assertions.assertEquals;
|
||||
import static org.junit.jupiter.api.Assertions.assertTimeout;
|
||||
|
||||
/**
|
||||
* test for Cluster Analysis.
|
||||
*
|
||||
* @author Julian Hinxlage
|
||||
* @version 0.1
|
||||
*/
|
||||
class ClusterTest {
|
||||
|
||||
@Test
|
||||
void test100() {
|
||||
URL url = getClass().getClassLoader().getResource("kd100.csv");
|
||||
assert url != null;
|
||||
|
||||
Cluster cluster = new Cluster(new File(url.getFile()));
|
||||
|
||||
AtomicReference<Map<Integer, Map<String, String>>> results = new AtomicReference<>();
|
||||
|
||||
|
||||
assertTimeout(Duration.ofMillis(2000), () -> results.set(cluster.getResult()));
|
||||
|
||||
assertEquals("m", results.get().get(0).get("Geschlecht"));
|
||||
}
|
||||
|
||||
@Test
|
||||
void test1000() {
|
||||
URL url = getClass().getClassLoader().getResource("kd1000.csv");
|
||||
assert url != null;
|
||||
|
||||
Cluster cluster = new Cluster(new File(url.getFile()));
|
||||
|
||||
AtomicReference<Map<Integer, Map<String, String>>> results = new AtomicReference<>();
|
||||
|
||||
assertTimeout(Duration.ofMillis(2000), () -> results.set(cluster.getResult()));
|
||||
|
||||
assertEquals("m", results.get().get(0).get("Geschlecht"));
|
||||
}
|
||||
|
||||
@Test
|
||||
void test10000() {
|
||||
URL url = getClass().getClassLoader().getResource("kd10000.csv");
|
||||
assert url != null;
|
||||
|
||||
Cluster cluster = new Cluster(new File(url.getFile()));
|
||||
|
||||
AtomicReference<Map<Integer, Map<String, String>>> results = new AtomicReference<>();
|
||||
|
||||
assertTimeout(Duration.ofMillis(2000), () -> results.set(cluster.getResult()));
|
||||
|
||||
assertEquals("m", results.get().get(0).get("Geschlecht"));
|
||||
}
|
||||
}
|
Loading…
Reference in New Issue