Skip to content

Commit 5b118b1

Browse files
kumarivinkumarivin
authored
commit
commit
1 parent 35ebb43 commit 5b118b1

File tree

1 file changed

+253
-0
lines changed

1 file changed

+253
-0
lines changed

‎Multiobjective optimization/App7.java

Lines changed: 253 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,253 @@
1+
package com.marketing.Project_Recommend;
2+
import java.sql.Date;
3+
import java.text.DateFormat;
4+
import java.text.SimpleDateFormat;
5+
import java.util.regex.Matcher;
6+
import java.util.regex.Pattern;
7+
import java.io.BufferedReader;
8+
import java.io.File;
9+
import java.io.FileReader;
10+
import java.io.FileWriter;
11+
import java.io.IOException;
12+
import java.util.ArrayList;
13+
import java.util.Arrays;
14+
import java.util.Collection;
15+
import java.util.Collections;
16+
import java.util.Comparator;
17+
import java.util.HashMap;
18+
import java.util.Iterator;
19+
import java.util.LinkedHashMap;
20+
import java.util.List;
21+
import java.util.Map;
22+
import java.util.Map.Entry;
23+
import java.util.NavigableSet;
24+
import java.util.PriorityQueue;
25+
import java.util.Properties;
26+
import java.util.Set;
27+
import java.util.TreeMap;
28+
29+
import org.apache.commons.collections.MapUtils;
30+
import org.apache.mahout.cf.taste.common.NoSuchUserException;
31+
import org.apache.mahout.cf.taste.impl.model.file.FileDataModel;
32+
import org.apache.mahout.cf.taste.impl.neighborhood.NearestNUserNeighborhood;
33+
import org.apache.mahout.cf.taste.impl.neighborhood.ThresholdUserNeighborhood;
34+
import org.apache.mahout.cf.taste.impl.recommender.GenericItemBasedRecommender;
35+
import org.apache.mahout.cf.taste.impl.recommender.GenericUserBasedRecommender;
36+
import org.apache.mahout.cf.taste.impl.similarity.PearsonCorrelationSimilarity;
37+
import org.apache.mahout.cf.taste.model.DataModel;
38+
import org.apache.mahout.cf.taste.neighborhood.UserNeighborhood;
39+
import org.apache.mahout.cf.taste.recommender.ItemBasedRecommender;
40+
import org.apache.mahout.cf.taste.recommender.RecommendedItem;
41+
import org.apache.mahout.cf.taste.recommender.Recommender;
42+
import org.apache.mahout.cf.taste.recommender.UserBasedRecommender;
43+
import org.apache.mahout.cf.taste.similarity.ItemSimilarity;
44+
import org.apache.mahout.cf.taste.similarity.UserSimilarity;
45+
46+
import com.google.common.collect.ArrayListMultimap;
47+
import com.google.common.collect.HashBasedTable;
48+
import com.google.common.collect.Iterables;
49+
import com.google.common.collect.ListMultimap;
50+
import com.google.common.collect.Multimap;
51+
import com.google.common.collect.SortedSetMultimap;
52+
import com.google.common.collect.Table;
53+
import com.google.common.collect.TreeMultimap;
54+
import com.google.common.primitives.Ints;
55+
56+
/**
57+
* Hello world!
58+
*
59+
*/
60+
public class App7
61+
{
62+
private static final int Float = 0;
63+
private static final int List = 0;
64+
65+
public static void main( String[] args ) throws Exception
66+
{
67+
for (int i=0; i<= 11;i++){
68+
String temp ="C:/Users/TG/Documents/Independent study/";
69+
DataModel model = new FileDataModel(new File(temp+"ratings_train"+Integer.toString(i)+"n.csv"));
70+
UserSimilarity similarity = new PearsonCorrelationSimilarity(model);
71+
ItemSimilarity isimilarity = new PearsonCorrelationSimilarity(model);
72+
//modified
73+
UserNeighborhood neighborhood = new NearestNUserNeighborhood(10, similarity, model);
74+
UserBasedRecommender recommender = new GenericUserBasedRecommender(model, neighborhood, similarity);
75+
/*
76+
UserNeighborhood neighborhood = new ThresholdUserNeighborhood(0.1, similarity, model);
77+
UserBasedRecommender recommender = new GenericUserBasedRecommender(model, neighborhood, similarity);
78+
*/
79+
ItemBasedRecommender irecommender = new GenericItemBasedRecommender(model,isimilarity);
80+
//Code to retreive top movies from training set
81+
BufferedReader traindatareader = new BufferedReader(new FileReader(temp+"ratings_train"+Integer.toString(i)+"n.csv"));
82+
Multimap<Long,Object> movie_ratings = ArrayListMultimap.create();
83+
String line1 = null;
84+
85+
while ((line1 = traindatareader.readLine()) != null) {
86+
String[] words1 = line1.split(",");
87+
if(line1.equals(""))
88+
continue;
89+
long arg01=Long.parseLong(words1[1]);
90+
float arg11=Integer.parseInt(words1[2]);
91+
if (movie_ratings.containsKey(arg01)){
92+
float temp_rating=(java.lang.Float)(Iterables.get(movie_ratings.get((long) arg01),0));
93+
temp_rating=(temp_rating+arg11);
94+
int temp_count= (Integer) Iterables.get(movie_ratings.get((long) arg01), 1);
95+
temp_count=temp_count+1;
96+
movie_ratings.removeAll(arg01);
97+
movie_ratings.put(arg01, temp_rating);
98+
movie_ratings.put(arg01, temp_count);
99+
100+
}
101+
else{
102+
movie_ratings.put(arg01,arg11);
103+
movie_ratings.put(arg01,1);
104+
}
105+
}
106+
107+
108+
//Retrieve movies with minimum view ability of 5
109+
TreeMultimap<Long,Float> filtered_movie_ratings = TreeMultimap.create();
110+
for( Long movie:movie_ratings.keySet() ){
111+
int temp1=(Integer) Iterables.get(movie_ratings.get((long) movie), 1);
112+
if (temp1 > 5){
113+
float temp2=(java.lang.Float) Iterables.get(movie_ratings.get((long) movie), 0);
114+
filtered_movie_ratings.put(movie,(float)temp2/temp1);
115+
filtered_movie_ratings.put(movie, (float) temp1);
116+
}
117+
118+
}
119+
//Pass the filtered movies for sorting
120+
List<Long> top10=topNKeys(filtered_movie_ratings, 10);
121+
//Read all the test file into a hashmap list for convinience of calculating precision and recall
122+
BufferedReader reader = new BufferedReader(new FileReader(temp+"ratings_test"+Integer.toString(i)+"n.csv"));
123+
Multimap<Long,Object> myMultimap = ArrayListMultimap.create();
124+
Table<Long, Long, Float> testTable = HashBasedTable.create();
125+
Multimap<Long,Object> User_Prec_rec = ArrayListMultimap.create();
126+
long hits=0;
127+
long miss=0;
128+
String line = null;
129+
while ((line = reader.readLine()) != null) {
130+
String[] words = line.split(",");
131+
if(line.equals(""))
132+
continue;
133+
long arg0=Long.parseLong(words[0]);
134+
long arg1=Long.parseLong(words[1]);
135+
float arg2=Integer.parseInt(words[2]);
136+
testTable.put(arg0, arg1, arg2);
137+
//if (arg0==5485){
138+
//System.out.println(arg0+" "+ arg1+" "+arg2);
139+
140+
//}
141+
}
142+
reader.close();
143+
//System.out.println("print "+(testTable.get((long)5485,(long)588)<3.0));
144+
//Retreive all users in test dataset for which we have to calculate precision
145+
Set<Long> userlist=testTable.rowKeySet();
146+
float total_relevant_items=(float) 0.0;
147+
float allhit=(float) 0.0;
148+
float totalusers_intrain=(float)0.0;
149+
float totalprec=(float)0.0;
150+
float totalrecommendations=(float) 0.0;
151+
float no_of_users_in_test=userlist.size();
152+
for(Long user: userlist){
153+
float prec=0;
154+
float rec=0;
155+
//Code to retrieved the actual movies that the user under consideration has seen and reviewed
156+
totalusers_intrain+=1.0;
157+
float userlen=(testTable.row((long)user).keySet().size()); //temp_user.size();
158+
159+
float count_hit=0;
160+
try{
161+
//Code to retrieve recommended movies and ratings for the user under consideration , only top 10
162+
java.util.List<RecommendedItem> recommendations = recommender.recommend(user, 10);
163+
164+
if (recommendations.size()>0){
165+
totalrecommendations+=recommendations.size();
166+
total_relevant_items+=userlen;
167+
for (RecommendedItem recommendation : recommendations) {
168+
169+
//Extract recommended item, verify if it exist in actual data for the given user
170+
long recommended_movie = recommendation.getItemID();
171+
172+
if (testTable.contains((long)user, (long)recommended_movie) && testTable.get((long)user, (long)recommended_movie)>3.0){
173+
count_hit+=1;
174+
allhit+=1;
175+
}
176+
}
177+
prec=(float)(count_hit/recommendations.size());
178+
totalprec+=prec;
179+
rec=count_hit/userlen;
180+
User_Prec_rec.put(user,(float)1.0);
181+
User_Prec_rec.put(user,prec);
182+
User_Prec_rec.put(user,rec);
183+
User_Prec_rec.put(user,count_hit);
184+
User_Prec_rec.put(user,userlen);
185+
hits+=1;
186+
}
187+
else{
188+
Iterator<Long> topmv = top10.iterator();
189+
totalrecommendations+=10;
190+
total_relevant_items+=userlen;
191+
while (topmv.hasNext()) {
192+
//Extract recommended item, verify if it exist in actual data for the given user
193+
long mv= topmv.next();
194+
if (testTable.contains((long)user,(long)mv) && testTable.get((long)user,(long)mv)>3.0){
195+
count_hit+=1;
196+
allhit+=1;
197+
}
198+
}
199+
prec=count_hit/10;
200+
totalprec+=prec;
201+
rec=count_hit/userlen;
202+
User_Prec_rec.put(user,(float)0.0);
203+
User_Prec_rec.put(user,prec);
204+
User_Prec_rec.put(user,rec);
205+
User_Prec_rec.put(user,count_hit);
206+
User_Prec_rec.put(user,userlen);
207+
miss+=1;
208+
}
209+
}
210+
catch(NoSuchUserException e){
211+
System.out.println("test");
212+
}
213+
}
214+
float total_precision= allhit/(totalrecommendations);
215+
//totalprec/totalusers_intrain;
216+
float total_recall=allhit/(total_relevant_items);
217+
try
218+
{
219+
String filename= "Data/precrec10_1.txt";
220+
FileWriter fw = new FileWriter(filename,true); //the true will append the new data
221+
fw.write("epoch :"+Integer.toString(i)+'\n');
222+
fw.write("hits: "+ Long.toString(hits)+ " miss: "+Long.toString(miss)+ " total relevant items in recommendations: "+Long.toString((long) allhit)+" total recommendations: "+Long.toString((long) totalrecommendations)+ " total relevant items: "+Long.toString((long) total_relevant_items)+'\n');
223+
fw.write(User_Prec_rec.toString()+'\n');
224+
fw.write("Total Precision: "+ (total_precision)+ " Total recall: "+(total_recall)+'\n');
225+
fw.close();
226+
}
227+
catch(IOException ioe){
228+
System.err.println("IOException: " + ioe.getMessage());
229+
}
230+
}
231+
}
232+
public static List<Long> topNKeys(final TreeMultimap<Long, Float> map, int n) {
233+
PriorityQueue<Long> topN = new PriorityQueue<Long>(n, new Comparator<Long>() {
234+
public int compare(Long o1, Long o2) {
235+
float o1r=(java.lang.Float) Iterables.get(map.get((long) o1), 0);
236+
float o2r=(java.lang.Float) Iterables.get(map.get((long) o2), 0);
237+
return Double.compare(o1r, o2r);
238+
}
239+
});
240+
241+
for(Long key:map.keySet()){
242+
if (topN.size() < n)
243+
topN.add(key);
244+
245+
else if (Iterables.get(map.get((long) topN.peek()), 0)< Iterables.get(map.get((long) key), 0)) {
246+
topN.poll();
247+
topN.add(key);
248+
}
249+
}
250+
return (List) Arrays.asList(topN.toArray());
251+
}
252+
253+
}

0 commit comments

Comments
 (0)