java - 尝试使用 Weka 将更多实例添加到训练集中时出现 IndexOutOfBoundsException

Question

我正在尝试向我的训练集添加更多实例并执行 10 倍交叉验证。

我的实例采用字符串格式，因此我使用 StringToWordVector 过滤器将它们转换为数字。如果我不添加我想要的额外页面，事情就会顺利进行。但是当我添加命令 trainSet.addAll(data2); 并将 trainSet 传递给过滤器时，我在第一次迭代中得到一个奇怪的 IndexOutOfBoundsException 实例fTrainSet = Filter.useFilter(trainSet, filter);

Instances data = getDataFromFile("pathtofile.arff");//main dataset 1821 instances
Instances data2 = getDataFromFile("anotherpath.arff");//709 instances i want to add 
int folds = 10;
for(int i=0;i<folds;i++){
    Instances trainSet = data.trainCV(folds, i);//training set
    System.out.println(trainSet.numInstances());//Prints 1638
    Instances testSet =  data.testCV(folds, i);//testing set

    //add more instances
    trainSet.addAll(data2);        
    System.out.println(trainSet.numInstances());//Prints 2347

    //filter
    StringToWordVector filter = new StringToWordVector();
    filter.setInputFormat(trainSet);        
    filter.setWordsToKeep(10000);
    filter.setTFTransform(true);
    filter.setLowerCaseTokens(true);
    filter.setOutputWordCounts(true);
    Stemmer stemmer = new IteratedLovinsStemmer();
    filter.setStemmer(stemmer);
    WordsFromFile stopwords = new WordsFromFile();
    stopwords.setStopwords(new File(".data/stopwords2.txt"));
    filter.setStopwordsHandler(stopwords);

    Instances fTrainSet = Filter.useFilter(trainSet, filter);//error!!!
    Instances fTestSet = Filter.useFilter(testSet, filter);
    ....
    //classification and evaluation....

当我尝试使用过滤器时出现以下错误:

Exception in thread "main" java.lang.IndexOutOfBoundsException: Index: 2161, Size: 1749
    at java.util.ArrayList.rangeCheck(Unknown Source)
    at java.util.ArrayList.get(Unknown Source)
    at weka.core.Attribute.addStringValue(Attribute.java:924)
    at weka.core.StringLocator.copyStringValues(StringLocator.java:150)
    at weka.core.StringLocator.copyStringValues(StringLocator.java:91)
    at weka.filters.Filter.copyValues(Filter.java:399)
    at weka.filters.Filter.bufferInput(Filter.java:342)
    at weka.filters.unsupervised.attribute.StringToWordVector.input(StringToWordVector.java:655)
    at weka.filters.Filter.useFilter(Filter.java:692)
    at CrossValidationExample.main(CrossValidationExample.java:108)

可能出了什么问题？

Answer 1

经过一番搜索，我意识到addAll有问题。功能。我能想到的一个原因是 addAll只是添加实例的引用，当我尝试将它们与 filter 一起使用时，这是一个问题。相反，我使用了此处建议的合并功能 https://stackoverflow.com/a/12359788/3923800 ，所以我替换了trainSet.addAll(data2);与 Instances newTrainSettrainSet = merge(trainSet,data2);一切正常。