An algorithm based on Java implementation, can automatically check the set of outliers in a set of data, eliminate these outliers, and finally get normal data.基于java实现的能够自动检查出一组数据中的异常值的集合，剔除这些异常集，得到正常数据。
可以这样，可不用考虑索引顺序影响：
Set set = oc(sensitivity, datas);
double[] doubles = new double[datas.length - set.size()];
int index = 0;
for (int i = 0; i < datas.length; i++) {
if (!set.contains(i)) {
doubles[index++] = datas[i];
}
}
return doubles;
shileiyuan86
commented
2 years ago
fusionToList方法存在bug，Set是无序集合，里面异常数据索引有可能不是按照正常顺序存放的。所以在for循环中不会与datas索引一样按顺序取数据。 @Override public double[] fusionToList(int sensitivity, double[] datas) { Set set = oc(sensitivity, datas);
double[] doubles = new double[datas.length - set.size()];
Iterator iterator = set.iterator();
Integer next = iterator.hasNext() ? iterator.next() : null;
int index = 0;
for (int i = 0; i < datas.length; i++) {
if (next != null && next.equals(i)) {
next = iterator.hasNext() ? iterator.next() : null;
continue;
}
doubles[index++] = datas[i];
}
return doubles;
}