in core/src/main/java/org/apache/sdap/mudrod/recommendation/structure/PODAACMetadataFeature.java [288:325]
public void categoricalVariablesSimilarity(Map<String, Object> metadataA, Map<String, Object> metadataB, XContentBuilder contentBuilder) {
// TODO Auto-generated method stub
for (String variable : featureTypes.keySet()) {
Integer type = featureTypes.get(variable);
if (!type.equals(VAR_CATEGORICAL)) {
continue;
}
double similarity = 0.0;
Object valueA = metadataA.get(variable);
Object valueB = metadataB.get(variable);
if (valueA instanceof ArrayList) {
ArrayList<String> aList = (ArrayList<String>) valueA;
ArrayList<String> bList = (ArrayList<String>) valueB;
if (aList != null && bList != null) {
int lengthA = aList.size();
int lengthB = bList.size();
List<String> newAList = new ArrayList<>(aList);
List<String> newBList = new ArrayList<>(bList);
newAList.retainAll(newBList);
similarity = (double)newAList.size() / lengthA;
}
} else if (valueA instanceof String) {
if (valueA.equals(valueB)) {
similarity = 1.0;
}
}
try {
contentBuilder.field(variable + "_Sim", similarity);
} catch (IOException e) {
// TODO Auto-generated catch block
LOG.error("Categorical variables similarity calculation failed : ", e);
}
}
}