我有一个数组，我想分成更小的n个数组，并对每个数组执行操作。我目前的做法是
用Java中的ArrayLists实现（任何伪代码都可以）
for (int i = 1; i <= Math.floor((A.size() / n)); i++) {
ArrayList temp = subArray(A, ((i * n) - n),
(i * n) - 1);
// do stuff with temp
}
private ArrayList<Comparable> subArray(ArrayList A, int start,
int end) {
ArrayList toReturn = new ArrayList();
for (int i = start; i <= end; i++) {
toReturn.add(A.get(i));
}
return toReturn;
}
其中A是列表，n是所需列表的大小
我相信这种方式在处理相当大的列表（大小高达100万）时花费了太多时间，所以我试图弄清楚什么会更有效率。
你想要做一些利用List.subList(int, int)视图而不是复制每个子列表的东西。要轻松地做到这一点，请使用Guava的Lists.partition(List, int)方法：
List<Foo> foos = ...
for (List<Foo> partition : Lists.partition(foos, n)) {
// do something with partition
}
请注意，与许多事物一样，这不是非常有效的List不是RandomAccess（如LinkedList）。
这是一种将List分区为子列表数组的方法，它确保除最后一个子列表之外的所有子列表都具有相同数量的元素：
static <T> List<T>[] split(List<T> source, int numPartitions) {
if (numPartitions < 2)
return new List[]{source};
final int sourceSize = source.size(),
partitions = numPartitions > sourceSize ? sourceSize: numPartitions,
increments = sourceSize / partitions;
return IntStream.rangeClosed(0, partitions)
.mapToObj(i -> source.subList(i*increments, Math.min((i+1)*increments, sourceSize)))
.toArray(List[]::new);
}
如果你想保证numPartitions数组大小，那么你想要：
static <T> List<T>[] split(List<T> source, int numPartitions) {
if (numPartitions < 2)
return new List[]{source};
final int sourceSize = source.size(),
partitions = numPartitions > sourceSize ? sourceSize: numPartitions,
increments = sourceSize / partitions;
return IntStream.range(0, partitions)
.mapToObj(i -> source.subList(i*increments, i == partitions-1 ? sourceSize : (i+1)*increments))
.toArray(List[]::new);
}
import java.util.ArrayList;
import java.util.Arrays;
import java.util.List;
public class SubListTest
{
public static void main(String[] args)
{
List<String> alphabetNames = new ArrayList<String>();
// populate alphabetNames array with AAA,BBB,CCC,.....
int a = (int) 'A';
for (int i = 0; i < 26; i++)
{
char x = (char) (a + i);
char[] array = new char[3];
Arrays.fill(array, x);
alphabetNames.add(new String(array));
}
int[] maxListSizes = new int[]
{
5, 10, 15, 20, 25, 30
};
for (int maxListSize : maxListSizes)
{
System.out.println("######################################################");
System.out.println("Partitioning original list of size " + alphabetNames.size() + " in to sub lists of max size "
+ maxListSize);
ArrayList<List<String>> subListArray = new ArrayList<List<String>>();
if (alphabetNames.size() <= maxListSize)
{
subListArray.add(alphabetNames);
}
else
{
// based on subLists of maxListSize X
int subListArraySize = (alphabetNames.size() + maxListSize - 1) / maxListSize;
for (int i = 0; i < subListArraySize; i++)
{
subListArray.add(alphabetNames.subList(i * maxListSize,
Math.min((i * maxListSize) + maxListSize, alphabetNames.size())));
}
}
System.out.println("Resulting number of partitions " + subListArray.size());
for (List<String> subList : subListArray)
{
System.out.println(subList);
}
}
}
}
输出：
######################################################
Partitioning original list of size 26 in to sub lists of max size 5
Resulting number of partitions 6
[AAA, BBB, CCC, DDD, EEE]
[FFF, GGG, HHH, III, JJJ]
[KKK, LLL, MMM, NNN, OOO]
[PPP, QQQ, RRR, SSS, TTT]
[UUU, VVV, WWW, XXX, YYY]
[ZZZ]
######################################################
Partitioning original list of size 26 in to sub lists of max size 10
Resulting number of partitions 3
[AAA, BBB, CCC, DDD, EEE, FFF, GGG, HHH, III, JJJ]
[KKK, LLL, MMM, NNN, OOO, PPP, QQQ, RRR, SSS, TTT]
[UUU, VVV, WWW, XXX, YYY, ZZZ]
######################################################
Partitioning original list of size 26 in to sub lists of max size 15
Resulting number of partitions 2
[AAA, BBB, CCC, DDD, EEE, FFF, GGG, HHH, III, JJJ, KKK, LLL, MMM, NNN, OOO]
[PPP, QQQ, RRR, SSS, TTT, UUU, VVV, WWW, XXX, YYY, ZZZ]
######################################################
Partitioning original list of size 26 in to sub lists of max size 20
Resulting number of partitions 2
[AAA, BBB, CCC, DDD, EEE, FFF, GGG, HHH, III, JJJ, KKK, LLL, MMM, NNN, OOO, PPP, QQQ, RRR, SSS, TTT]
[UUU, VVV, WWW, XXX, YYY, ZZZ]
######################################################
Partitioning original list of size 26 in to sub lists of max size 25
Resulting number of partitions 2
[AAA, BBB, CCC, DDD, EEE, FFF, GGG, HHH, III, JJJ, KKK, LLL, MMM, NNN, OOO, PPP, QQQ, RRR, SSS, TTT, UUU, VVV, WWW, XXX, YYY]
[ZZZ]
######################################################
Partitioning original list of size 26 in to sub lists of max size 30
Resulting number of partitions 1
[AAA, BBB, CCC, DDD, EEE, FFF, GGG, HHH, III, JJJ, KKK, LLL, MMM, NNN, OOO, PPP, QQQ, RRR, SSS, TTT, UUU, VVV, WWW, XXX, YYY, ZZZ]
关于什么
Arrays.copyOfRange( original, from, to )
?
使用Java 8的一行：
IntStream.range(0, list.size() / batchSize + 1)
.mapToObj(i -> list.subList(i * batchSize,
Math.min(i * batchSize + batchSize, list.size())))
.filter(s -> !s.isEmpty()).collect(Collectors.toList());
如果您正在使用列表，我使用“Apache Commons Collections 4”库。它在ListUtils类中有一个分区方法：
...
int targetSize = 100;
List<Integer> largeList = ...
List<List<Integer>> output = ListUtils.partition(largeList, targetSize);
例如：
int partitionSize = 10;
List<List<String>> partitions = new ArrayList<>();
for (int i=0; i<yourlist.size(); i += partitionSize) {
partitions.add(yourlist.subList(i, Math.min(i + partitionSize, yourlist.size())));
}
for (List<String> list : partitions) {
//Do your stuff on each sub list
}
好吧，在我看到ColinD的回答（+1）之前我自己写了一篇，使用Guava绝对是要走的路。单独留下太有趣了，所以下面给出了列表的副本而不是视图，因此GUava的效率肯定比这更高。我发布这个是因为写它很有趣而不是暗示它有效：
Hamcrest测试（无论如何）：
assertThat(chunk(asList("a", "b", "c", "d", "e"), 2),
equalTo(asList(asList("a", "b"), asList("c", "d"), asList("e"))));
代码：
public static <T> Iterable<Iterable<T>> chunk(Iterable<T> in, int size) {
List<Iterable<T>> lists = newArrayList();
Iterator<T> i = in.iterator();
while (i.hasNext()) {
List<T> list = newArrayList();
for (int j=0; i.hasNext() && j<size; j++) {
list.add(i.next());
}
lists.add(list);
}
return lists;
}
public <E> Iterable<List<E>> partition(List<E> list, final int batchSize)
{
assert(batchSize > 0);
assert(list != null);
assert(list.size() + batchSize <= Integer.MAX_VALUE); //avoid overflow
int idx = 0;
List<List<E>> result = new ArrayList<List<E>>();
for (idx = 0; idx + batchSize <= list.size(); idx += batchSize) {
result.add(list.subList(idx, idx + batchSize));
}
if (idx < list.size()) {
result.add(list.subList(idx, list.size()));
}
return result;
}
如果您不想使用库，这是我的解决方案
1.分为N个相等的部分：
private <T> List<List<T>> nPartition(List<T> objs, final int N) {
return new ArrayList<>(IntStream.range(0, objs.size()).boxed().collect(
Collectors.groupingBy(e->e%N,Collectors.mapping(e->objs.get(e), Collectors.toList())
)).values());
}
2.要分组N个项目：
private <T> List<List<T>> nPartition(List<T> objs, final int N) {
return new ArrayList<>(IntStream.range(0, objs.size()).boxed().collect(
Collectors.groupingBy(e->e/N,Collectors.mapping(e->objs.get(e), Collectors.toList())
)).values());
}
我刚刚实现了一个列表分区，因为我无法使用库。
所以我想在这里分享我的代码：
import java.util.Iterator;
import java.util.List;
import java.util.NoSuchElementException;
public class ListPartitioning<T> implements Iterable<List<T>> {
private final List<T> list;
private final int partitionSize;
public ListPartitioning(List<T> list, int partitionSize) {
if (list == null) {
throw new IllegalArgumentException("list must not be null");
}
if (partitionSize < 1) {
throw new IllegalArgumentException("partitionSize must be 1 or greater");
}
this.list = list;
this.partitionSize = partitionSize;
}
@Override
public Iterator<List<T>> iterator() {
return new ListPartitionIterator<T>(list, partitionSize);
}
private static class ListPartitionIterator<T> implements Iterator<List<T>> {
private int index = 0;
private List<T> listToPartition;
private int partitionSize;
private List<T> nextPartition;
public ListPartitionIterator(List<T> listToPartition, int partitionSize) {
this.listToPartition = listToPartition;
this.partitionSize = partitionSize;
}
@Override
public boolean hasNext() {
return index < listToPartition.size();
}
@Override
public List<T> next() {
if (!hasNext()) {
throw new NoSuchElementException();
}
int partitionStart = index;
int partitionEnd = Math.min(index + partitionSize, listToPartition.size());
nextPartition = listToPartition.subList(partitionStart, partitionEnd);
index = partitionEnd;
return nextPartition;
}
@Override
public void remove() {
if (nextPartition == null) {
throw new IllegalStateException("next must be called first");
}
nextPartition.clear();
index -= partitionSize;
nextPartition = null;
}
}
}
并且基于testng的单元测试。
import org.testng.Assert;
import org.testng.annotations.Test;
import java.util.*;
public class ListPartitioningTest {
@Test(expectedExceptions = IllegalArgumentException.class)
public void nullList() {
ListPartitioning<String> lists = new ListPartitioning<String>(null, 1);
}
@Test(groups = Group.UNIT_TEST, expectedExceptions = IllegalArgumentException.class)
public void wrongPartitionSize() {
ListPartitioning<String> lists = new ListPartitioning<String>(new ArrayList<String>(), 0);
}
@Test()
public void iteratorTest() {
List<Integer> integers = Arrays.asList(0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15);
ListPartitioning<Integer> listPartitioning = new ListPartitioning<Integer>(integers, 7);
Iterator<List<Integer>> partitionIterator = listPartitioning.iterator();
Assert.assertNotNull(partitionIterator);
Assert.assertTrue(partitionIterator.hasNext(), "next partition (first)");
List<Integer> partition = partitionIterator.next();
Assert.assertEquals(partition, Arrays.asList(0, 1, 2, 3, 4, 5, 6));
Assert.assertTrue(partitionIterator.hasNext(), "next partition (second)");
partition = partitionIterator.next();
Assert.assertEquals(partition, Arrays.asList(7, 8, 9, 10, 11, 12, 13));
Assert.assertTrue(partitionIterator.hasNext(), "next partition (third)");
partition = partitionIterator.next();
Assert.assertEquals(partition, Arrays.asList(14, 15));
Assert.assertFalse(partitionIterator.hasNext());
}
@Test(expectedExceptions = NoSuchElementException.class)
public void noSuchElementException() {
List<Integer> integers = Arrays.asList(1);
ListPartitioning<Integer> listPartitioning = new ListPartitioning<Integer>(integers, 2);
Iterator<List<Integer>> partitionIterator = listPartitioning.iterator();
List<Integer> partition = partitionIterator.next();
partition = partitionIterator.next();
}
@Test(expectedExceptions = IllegalStateException.class)
public void removeWithoutNext() {
List<Integer> integers = new ArrayList<Integer>(Arrays.asList(0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15));
ListPartitioning<Integer> listPartitioning = new ListPartitioning<Integer>(integers, 7);
Iterator<List<Integer>> partitionIterator = listPartitioning.iterator();
partitionIterator.remove();
}
@Test()
public void remove() {
List<Integer> integers = new ArrayList<Integer>(Arrays.asList(0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15));
ListPartitioning<Integer> listPartitioning = new ListPartitioning<Integer>(integers, 7);
Iterator<List<Integer>> partitionIterator = listPartitioning.iterator();
partitionIterator.next();
partitionIterator.next();
partitionIterator.remove();
Assert.assertTrue(partitionIterator.hasNext(), "next partition ");
List<Integer> partition = partitionIterator.next();
Assert.assertEquals(partition, Arrays.asList(14, 15));
Assert.assertFalse(partitionIterator.hasNext());
Assert.assertEquals(integers, Arrays.asList(0, 1, 2, 3, 4, 5, 6, 14, 15));
}
}
如果您正在处理数组，可以使用System.arraycopy()。
int[] a = {1,2,3,4,5};
int[] b = new int[2];
int[] c = new int[3];
System.arraycopy(a, 0, b, 0, 2); // b will be {1,2}
System.arraycopy(a, 2, c, 0, 3); // c will be {3,4,5}
由于您希望优化性能，因此应使用并行流而不是for循环。这样您就可以使用多个线程。
Lists.partition(A, n).parallelStream().forEach({
//do stuff with temp
});
您还可以使用其他方式与流进行混合，例如，如果匹配您的目的，则收集或映射。