Need Help for using Sparse Matrix

[sososa1213]

Here the code I use it and it works fine. However, I need to use sparse matrix for the dataset ml-100k->ratings.txt and store the result in new file. Really I get stuck in this point and I did not find any example for how use sparse matrix with new version of librec.

I really highly appreciate any help. Thank.

import net.librec.BaseTestCase;
import net.librec.common.LibrecException;
import net.librec.conf.Configuration;
import net.librec.conf.Configuration.Resource;
import net.librec.data.model.TextDataModel;
import net.librec.eval.RecommenderEvaluator;
import net.librec.eval.rating.MAEEvaluator;
import net.librec.eval.rating.RMSEEvaluator;
import net.librec.filter.GenericRecommendedFilter;
import net.librec.filter.RecommendedFilter;
import net.librec.recommender.Recommender;
import net.librec.recommender.RecommenderContext;
import net.librec.recommender.cf.ItemKNNRecommender;
import net.librec.recommender.cf.UserKNNRecommender;
import net.librec.recommender.item.RecommendedItem;
import net.librec.similarity.PCCSimilarity;
import net.librec.similarity.RecommenderSimilarity;
import java.awt.List;
import java.io.BufferedWriter;
import java.io.FileWriter;

import org.junit.Before;
import org.junit.FixMethodOrder;
import org.junit.Ignore;
import org.junit.Test;
import org.junit.runners.MethodSorters;
public class test1 {
    public static void main (String [] args)  throws Exception {
    System.out.print("xxx");
    Resource resource = new Resource("librec.properties");

    Configuration conf = new Configuration();
    conf.addResource(resource);
    conf.set("data.splitter.trainset.ratio", "0.8");

    conf.set("inputDataPath", conf.get("dfs.data.dir") + "/movielens/ml-100k/ratings.txt");
    conf.set("data.column.format","UIR");
    conf.set("data.convert.binarize.threshold","-1.0");

    TextDataModel dataModel = new TextDataModel(conf);
    dataModel.buildDataModel();

    RecommenderContext context = new RecommenderContext(conf,dataModel);

    RecommenderSimilarity similarity = new PCCSimilarity();
    similarity.buildSimilarityMatrix(dataModel);
    context.setSimilarity(similarity);

    Recommender recommender = new ItemKNNRecommender();
    recommender.setContext(context);

    recommender.recommend(context);

    RecommenderEvaluator evaluator = new MAEEvaluator();

    System.out.println("MAE:" + recommender.evaluate(evaluator));
    java.util.List<RecommendedItem> recommendedItemList = recommender.getRecommendedList();

//    print filter result
    System.out.println("len " + recommendedItemList.size());
    BufferedWriter writer = new BufferedWriter(new FileWriter("recommendations.txt"));
    for (RecommendedItem recommendedItem : recommendedItemList) {
        //System.out.println("user:" + recommendedItem.getUserId() + " " +
            //  "item:" + recommendedItem.getItemId() + " " +
            //  "value:" + recommendedItem.getValue() + "\n");
        writer.write(

                "user:" + recommendedItem.getUserId() + " " +
                "item:" + recommendedItem.getItemId() + " " +
                "value:" + recommendedItem.getValue() + "\n"
                );
        writer.flush();
    }
    writer.close();
    }
}

[allenjack]

Hi, Sorry, I did not get your point. Basically, the raw data is stored in the 'dataModel' object. We transform the raw data into the sparse matrix and store it in the 'recommender' object. There are trainMatrix and testMatrix in the 'recommender' object, which all are sparse matrices.

[sososa1213]

Thanks for your comment, I get confused do you mean I have to do this:

We transform the raw data into the sparse matrix and store it in the 'recommender' object.

or librec did it already?

Can you pleas check issue #266

[allenjack]

Hi, You do not need to do that. It has been done default by LibRec.

On 17 July 2018 at 03:58, sososa1213 notifications@github.com wrote:

Thanks for your comment, I get confused do you mean I have to do this:

We transform the raw data into the sparse matrix and store it in the 'recommender' object.

or librec did it already?

Can you pleas check issue #266 https://github.com/guoguibing/librec/issues/266

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/guoguibing/librec/issues/263#issuecomment-405494496, or mute the thread https://github.com/notifications/unsubscribe-auth/AEN9vss7s9Yw0JDLnMNo1QJrGp5UUqLMks5uHZkQgaJpZM4VO9kT .

[sososa1213]

Is that mean if I use the way I poste it in issue #266, that means I used sparse matrix?

Sorry but I wnat to make sure before finalizing my code, becasue I need to improve my code by using sparse matrix.

Thanks again @allenjack

[allenjack]

Yes, you are right. You are using the sparse matrix.

On 17 July 2018 at 10:44, sososa1213 notifications@github.com wrote:

Is that mean if I use the way I poste it in issue #266 https://github.com/guoguibing/librec/issues/266, that means I used sparse matrix?

Sorry but I wnat to make sure before finalizing my code, becasue I need to improve my code by using sparse matrix.

Thanks again @allenjack https://github.com/allenjack

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/guoguibing/librec/issues/263#issuecomment-405607756, or mute the thread https://github.com/notifications/unsubscribe-auth/AEN9vmFBcprwR4jyz69-1O02ZHUVhnyfks5uHfhVgaJpZM4VO9kT .

[sososa1213]

Thank you I really appreciat your help