View a markdown version of this page

$topN - Amazon DocumentDB
示例（MongoDB Shell）窗口运算符用法示例（MongoDB Shell）代码示例

本文属于机器翻译版本。若本译文内容与英语原文存在差异，则一律以英文原文为准。

$topN

8.0.1 版中的新增内容。

使用$group阶段中的$topN累加器根据指定的排序顺序返回组中前 N 个元素。如果一个组包含少于 N 个元素，则$topN返回该组中的所有元素。

参数

  • n：正整数或解析为一的表达式，指定每组要返回多少个顶级结果。

  • sortBy：指定排序顺序的文档。1用于升序或降序-1。

  • output：一个表达式，它指定要从前 N 个文档中返回的字段。

示例（MongoDB Shell）

以下示例说明如何使用$topN累加器来查找销售集合中每件商品的前 2 个销售额（最高数量）。

创建示例文档 

db.sales.insertMany([
  { item: "abc", quantity: 10, price: 5 },
  { item: "abc", quantity: 7, price: 8 },
  { item: "abc", quantity: 5, price: 10 },
  { item: "xyz", quantity: 15, price: 3 },
  { item: "xyz", quantity: 9, price: 6 },
  { item: "xyz", quantity: 3, price: 12 }
])

查询示例 

db.sales.aggregate([
  { $group: { _id: "$item", topTwoSales: { $topN: { n: 2, sortBy: { quantity: -1 }, output: { quantity: "$quantity", price: "$price" } } } } }
])

输出

[
  { "_id": "xyz", "topTwoSales": [{ "quantity": 15, "price": 3 }, { "quantity": 9, "price": 6 }] },
  { "_id": "abc", "topTwoSales": [{ "quantity": 10, "price": 5 }, { "quantity": 7, "price": 8 }] }
]

窗口运算符用法示例（MongoDB Shell）

8.0.2 版中的新内容。

该$topN操作员还可以在$setWindowFields舞台中用作窗口操作员。在这种情况下，它返回每个窗口中前 N 个文档（根据操作员自己的sortBy）中的一个数组。output您可以在output字段下指定运算符，也可以选择使用window文档定义窗口边界。

注意

在中用作窗口运算符时$setWindowFields，中间数据限制$topN为 100 MB。超过此限制的操作会返回错误。

创建示例文档 

db.matchScores.insertMany([
  { _id: 1, player: "Alice", round: 1, score: 20 },
  { _id: 2, player: "Alice", round: 2, score: 35 },
  { _id: 3, player: "Alice", round: 3, score: 28 },
  { _id: 4, player: "Bob", round: 1, score: 15 },
  { _id: 5, player: "Bob", round: 2, score: 40 }
]);

查询示例

以下示例对文档进行分区player，对每个分区进行排序round，并返回从分区开始到当前文档的两轮得分最高的回合。

db.matchScores.aggregate([
  {
    $setWindowFields: {
      partitionBy: "$player",
      sortBy: { round: 1 },
      output: {
        topTwoRoundsSoFar: {
          $topN: { n: 2, sortBy: { score: -1 }, output: { round: "$round", score: "$score" } },
          window: { documents: ["unbounded", "current"] }
        }
      }
    }
  }
]);

输出

[
  { "_id": 1, "player": "Alice", "round": 1, "score": 20, "topTwoRoundsSoFar": [{ "round": 1, "score": 20 }] },
  { "_id": 2, "player": "Alice", "round": 2, "score": 35, "topTwoRoundsSoFar": [{ "round": 2, "score": 35 }, { "round": 1, "score": 20 }] },
  { "_id": 3, "player": "Alice", "round": 3, "score": 28, "topTwoRoundsSoFar": [{ "round": 2, "score": 35 }, { "round": 3, "score": 28 }] },
  { "_id": 4, "player": "Bob", "round": 1, "score": 15, "topTwoRoundsSoFar": [{ "round": 1, "score": 15 }] },
  { "_id": 5, "player": "Bob", "round": 2, "score": 40, "topTwoRoundsSoFar": [{ "round": 2, "score": 40 }, { "round": 1, "score": 15 }] }
]

每个文档都添加了其topTwoRoundsSoFar分区内得分最高的两轮回合，直至包括当前文档。如果窗口包含的文档少于两个，则将返回所有文档。

代码示例

要查看使用$topN运算符的代码示例，请选择要使用的语言的选项卡。以下示例显示了累加器的用法（输入$group）和窗口运算符的用法（输入$setWindowFields）：

Node.js
const { MongoClient } = require('mongodb');

async function example() {
  const uri = 'mongodb://<username>:<password>@<cluster-endpoint>:27017/?tls=true&tlsCAFile=global-bundle.pem&replicaSet=rs0&readPreference=secondaryPreferred&retryWrites=false';
  const client = new MongoClient(uri);

  try {
    await client.connect();

    const db = client.db('test');

    // Accumulator usage: top N documents per group
    const sales = db.collection('sales');
    await sales.insertMany([
      { item: "abc", quantity: 10, price: 5 },
      { item: "abc", quantity: 7, price: 8 },
      { item: "abc", quantity: 5, price: 10 },
      { item: "xyz", quantity: 15, price: 3 },
      { item: "xyz", quantity: 9, price: 6 },
      { item: "xyz", quantity: 3, price: 12 }
    ]);
    const accumulatorResult = await sales.aggregate([
      { $group: { _id: "$item", topTwoSales: { $topN: { n: 2, sortBy: { quantity: -1 }, output: { quantity: "$quantity", price: "$price" } } } } }
    ]).toArray();
    console.log('Accumulator result:', accumulatorResult);

    // Window operator usage: top two rounds so far within each partition
    const matchScores = db.collection('matchScores');
    await matchScores.insertMany([
      { _id: 1, player: "Alice", round: 1, score: 20 },
      { _id: 2, player: "Alice", round: 2, score: 35 },
      { _id: 3, player: "Alice", round: 3, score: 28 },
      { _id: 4, player: "Bob", round: 1, score: 15 },
      { _id: 5, player: "Bob", round: 2, score: 40 }
    ]);
    const windowResult = await matchScores.aggregate([
      {
        $setWindowFields: {
          partitionBy: "$player",
          sortBy: { round: 1 },
          output: {
            topTwoRoundsSoFar: {
              $topN: { n: 2, sortBy: { score: -1 }, output: { round: "$round", score: "$score" } },
              window: { documents: ["unbounded", "current"] }
            }
          }
        }
      }
    ]).toArray();
    console.log('Window result:', windowResult);

  } catch (error) {
    console.error('Error:', error);
  } finally {
    await client.close();
  }
}

example();
Python
from pymongo import MongoClient
from pprint import pprint

def example():
    client = None
    try:
        client = MongoClient('mongodb://<username>:<password>@<cluster-endpoint>:27017/?tls=true&tlsCAFile=global-bundle.pem&replicaSet=rs0&readPreference=secondaryPreferred&retryWrites=false')

        db = client['test']

        # Accumulator usage: top N documents per group
        sales = db['sales']
        sales.insert_many([
            { 'item': 'abc', 'quantity': 10, 'price': 5 },
            { 'item': 'abc', 'quantity': 7, 'price': 8 },
            { 'item': 'abc', 'quantity': 5, 'price': 10 },
            { 'item': 'xyz', 'quantity': 15, 'price': 3 },
            { 'item': 'xyz', 'quantity': 9, 'price': 6 },
            { 'item': 'xyz', 'quantity': 3, 'price': 12 }
        ])
        accumulator_result = list(sales.aggregate([
            { '$group': { '_id': '$item', 'topTwoSales': { '$topN': { 'n': 2, 'sortBy': { 'quantity': -1 }, 'output': { 'quantity': '$quantity', 'price': '$price' } } } } }
        ]))
        pprint(accumulator_result)

        # Window operator usage: top two rounds so far within each partition
        match_scores = db['matchScores']
        match_scores.insert_many([
            { '_id': 1, 'player': 'Alice', 'round': 1, 'score': 20 },
            { '_id': 2, 'player': 'Alice', 'round': 2, 'score': 35 },
            { '_id': 3, 'player': 'Alice', 'round': 3, 'score': 28 },
            { '_id': 4, 'player': 'Bob', 'round': 1, 'score': 15 },
            { '_id': 5, 'player': 'Bob', 'round': 2, 'score': 40 }
        ])
        window_result = list(match_scores.aggregate([
            {
                '$setWindowFields': {
                    'partitionBy': '$player',
                    'sortBy': { 'round': 1 },
                    'output': {
                        'topTwoRoundsSoFar': {
                            '$topN': { 'n': 2, 'sortBy': { 'score': -1 }, 'output': { 'round': '$round', 'score': '$score' } },
                            'window': { 'documents': ['unbounded', 'current'] }
                        }
                    }
                }
            }
        ]))
        pprint(window_result)

    except Exception as e:
        print(f"An error occurred: {e}")

    finally:
        if client:
            client.close()

example()