// Determine F(k+1) by support counting on (C(K+1), T) and retaining itemsets from C(k+1) with
// support at least minsup
private static List<Itemset> determineFrequentItemsets(
List<Itemset> candicates, List<Transaction> transactions, double minsup) {
if (candicates.isEmpty()) {
return null;
}
HashTree hashTree = new HashTree(candicates, candicates.get(0).getNumOfItems());
HashMap<Itemset, Integer> frequentCount = new HashMap<>();
for (Itemset itemset : candicates) {
frequentCount.put(itemset, 0);
}
for (Transaction transaction : transactions) {
Set<Itemset> candidatesInTranscation = hashTree.candidatesInTransaction(transaction);
if (candidatesInTranscation == null) {
continue;
}
for (Itemset itemset : candidatesInTranscation) {
if (transaction.containItemset(itemset)) {
frequentCount.put(itemset, frequentCount.get(itemset) + 1);
}
}
}
List<Itemset> result = new ArrayList<>();
for (Itemset itemset : candicates) {
if ((double) (frequentCount.get(itemset)) / transactions.size() >= minsup) {
result.add(itemset);
}
}
return result;
}
public static void main(String[] args) {
// Minimum support value
double minsup = 0.144;
// double minsup = 0.04;
// Load data, you should set your own data file location here
File data =
new File("/Users/walker/Desktop/Courses/DataMining/Assignments/2/assignment2-data.txt");
ArrayList<Transaction> transactions = new ArrayList<>();
ArrayList<String> items = new ArrayList<>();
try (BufferedReader br =
new BufferedReader(new InputStreamReader(new FileInputStream(data), "utf-8"))) {
String line;
// Load item's name
if ((line = br.readLine()) != null) {
Collections.addAll(items, line.split("\\s+"));
}
// Load transactions
Transaction transaction;
while ((line = br.readLine()) != null) {
transaction = new Transaction(line);
transactions.add(transaction);
}
} catch (IOException e) {
e.printStackTrace();
}
List<Itemset> frequentItemset = apriori(transactions, minsup);
HashMap<Itemset, Integer> frequentCount = new HashMap<>();
for (Itemset itemset : frequentItemset) {
frequentCount.put(itemset, 0);
}
for (Transaction transaction : transactions) {
for (Itemset itemset : frequentItemset) {
if (transaction.containItemset(itemset)) {
frequentCount.put(itemset, frequentCount.get(itemset) + 1);
}
}
}
for (Itemset itemset : frequentItemset) {
System.out.print(itemset);
System.out.println((double) frequentCount.get(itemset) / transactions.size());
}
}
