From 8040f53a1e176a874d944f8c7688f1f28a22c089 Mon Sep 17 00:00:00 2001
From: Fengchao <fyuab@connect.ust.hk>
Date: Tue, 30 Jan 2018 15:45:14 +0800
Subject: [PATCH] Considering overlapped theoretical fragment ion multiple
 times.

This is what Comet and Xolik do.
---
 pom.xml                                       |  5 +++
 .../proteomics/Types/SparseBooleanVector.java | 38 +++++--------------
 .../Types/SparseBooleanVectorTest.java        | 30 ++-------------
 3 files changed, 19 insertions(+), 54 deletions(-)

diff --git a/pom.xml b/pom.xml
index a91e7ea..363fa12 100644
--- a/pom.xml
+++ b/pom.xml
@@ -66,6 +66,11 @@
             <artifactId>commons-math3</artifactId>
             <version>3.6.1</version>
         </dependency>
+        <dependency>
+            <groupId>com.google.guava</groupId>
+            <artifactId>guava</artifactId>
+            <version>23.6-jre</version>
+        </dependency>
     </dependencies>
 
     <repositories>
diff --git a/src/main/java/proteomics/Types/SparseBooleanVector.java b/src/main/java/proteomics/Types/SparseBooleanVector.java
index b8cf7ff..0485551 100644
--- a/src/main/java/proteomics/Types/SparseBooleanVector.java
+++ b/src/main/java/proteomics/Types/SparseBooleanVector.java
@@ -1,14 +1,15 @@
 package proteomics.Types;
 
-import java.util.HashSet;
-import java.util.Map;
+import com.google.common.collect.HashMultiset;
+import com.google.common.collect.Multiset;
+
 import java.util.Set;
 
 public class SparseBooleanVector {
 
-    private Set<Integer> sparse_vector = new HashSet<>();
+    private Multiset<Integer> sparse_vector = HashMultiset.create();
 
-    SparseBooleanVector(Set<Integer> sparse_vector) {
+    SparseBooleanVector(Multiset<Integer> sparse_vector) {
         this.sparse_vector = sparse_vector;
     }
 
@@ -18,43 +19,24 @@ public void put(int idx) {
         sparse_vector.add(idx);
     }
 
-    double norm2square() {
-        return sparse_vector.size();
-    }
-
-    public double dot(SparseVector other) { // caution: it will change the original Sparse Boolean Vector.
+    public double dot(SparseVector other) {
         double output = 0;
-        sparse_vector.retainAll(other.getIdxSet());
-        for (int i : sparse_vector) {
+        for (Integer i : sparse_vector) {
             output += other.get(i);
         }
         return output;
     }
 
-    double dot(SparseBooleanVector other) {
-        Set<Integer> intersectedKeys = new HashSet<>(sparse_vector);
-        intersectedKeys.retainAll(other.sparse_vector);
-        return intersectedKeys.size();
-    }
-
-    public int size() {
-        return sparse_vector.size();
-    }
-
-    public boolean contains(int v) {
-        return sparse_vector.contains(v);
-    }
-
     public Set<Integer> getIdxSet() {
-        return sparse_vector;
+        return sparse_vector.elementSet();
     }
 
     public boolean equals(Object other) {
         if (other instanceof SparseBooleanVector) {
             SparseBooleanVector temp = (SparseBooleanVector) other;
-            if (temp.size() == sparse_vector.size()) {
+            if (temp.sparse_vector.size() == sparse_vector.size()) {
                 for (int v : sparse_vector) {
-                    if (!temp.contains(v)) {
+                    if (!temp.sparse_vector.contains(v)) {
                         return false;
                     }
                 }
diff --git a/src/test/java/proteomics/Types/SparseBooleanVectorTest.java b/src/test/java/proteomics/Types/SparseBooleanVectorTest.java
index b551a50..bdc0dbc 100644
--- a/src/test/java/proteomics/Types/SparseBooleanVectorTest.java
+++ b/src/test/java/proteomics/Types/SparseBooleanVectorTest.java
@@ -1,5 +1,7 @@
 package proteomics.Types;
 
+import com.google.common.collect.HashMultiset;
+import com.google.common.collect.Multiset;
 import org.junit.Before;
 import org.junit.Test;
 
@@ -14,7 +16,7 @@ public class SparseBooleanVectorTest {
 
     @Before
     public void setUp() throws Exception {
-        Set<Integer> temp = new HashSet<>();
+        Multiset<Integer> temp = HashMultiset.create();
         temp.add(1);
         temp.add(3);
         temp.add(23);
@@ -23,7 +25,7 @@ public void setUp() throws Exception {
 
     @Test
     public void put() throws Exception {
-        Set<Integer> temp = new HashSet<>();
+        Multiset<Integer> temp = HashMultiset.create();
         temp.add(1);
         temp.add(3);
         temp.add(23);
@@ -34,11 +36,6 @@ public void put() throws Exception {
         assertEquals(ground_truth, vector);
     }
 
-    @Test
-    public void norm2square() throws Exception {
-        assertEquals(3, vector.norm2square(), 0);
-    }
-
     @Test
     public void dot() throws Exception {
         SparseVector other = new SparseVector();
@@ -47,25 +44,6 @@ public void dot() throws Exception {
         assertEquals(1.5f, vector.dot(other), 0);
     }
 
-    @Test
-    public void dot1() throws Exception {
-        SparseBooleanVector other = new SparseBooleanVector();
-        other.put(1);
-        other.put(2);
-        assertEquals(1, vector.dot(other), 0);
-    }
-
-    @Test
-    public void size() throws Exception {
-        assertEquals(3, vector.size());
-    }
-
-    @Test
-    public void contains() throws Exception {
-        assertTrue(vector.contains(1));
-        assertFalse(vector.contains(2));
-    }
-
     @Test
     public void equals() throws Exception {
         SparseBooleanVector other_1 = new SparseBooleanVector();