js使用groupby

前言

sql语句和pyhton中都有groupby的操作,处理数据来非常方便。而js对group不支持
于是想着自己写一个

单个字段group

单个字段聚合的话非常简单,


const arr = [
                {name: '小孙', age: 18, score: 60, weight: 60},
                {name: '小王', age: 19, score: 70, weight: 55},
                {name: '小李', age: 18, score: 60, weight: 70},
                {name: '小刘', age: 20, score: 70, weight: 65},
                {name: '小赵', age: 18, score: 60, weight: 60},
                {name: '小钱', age: 19, score: 70, weight: 55},
                {name: '小周', age: 20, score: 60, weight: 50},
        ];
const group_signal = (data, key) => {
            return data.reduce(function(prev, cur) {
                (prev[cur[key]] = prev[cur[key]] || []).push(cur);
                return prev;
            }, {});
        };
console.log(group_signal(arr, 'age'));
object: {18: Array(3), 19: Array(2), 20: Array(2)}
  18: Array(3)
    0: {name: "小孙", age: 18, score: 60, weight: 60}
    1: {name: "小李", age: 18, score: 60, weight: 70}
    2: {name: "小赵", age: 18, score: 60, weight: 60}
  19: Array(2)
    0: {name: "小王", age: 19, score: 70, weight: 55}
    1: {name: "小钱", age: 19, score: 70, weight: 55}
  20: Array(2)
    0: {name: "小刘", age: 20, score: 70, weight: 65}
    1: {name: "小周", age: 20, score: 60, weight: 50}


多个字段group

不光有单个字段聚合,也有多个字段聚合。比如我想找出 age 和 score 相同的。
多个字段聚合相对复杂些,首先要找到要聚合的字段,并且去掉重复的值。这里用到的函数比较多,我们将它封装成一个类

class GroupMultiple {
    //data是聚合的数据, name是聚合的字段
    constructor (data, name){
        this.data = data;
        this.name = name;
    }
}

然后定义一个getKeys方法取出聚合后的下标。例如按照 age 和 score 聚合 将 age 和 score 不同的值取出来放入一个数组。

class GroupMultiple {
    //data是聚合的数据, name是聚合的字段
    constructor (data, name){
        this.data = data;
        this.name = name;
    }
        
    getKeys(data, names){
        let _this = this;
        return data.reduce(function (prev, cur) {
            //首先拿到name对应的值, 将key组成 {key: value}形式
            var key = _this.getObject(cur, names);
            //判断数组里是否已经含有了key,有的话就跳过,否则把key放入到数组里
            //把判断方法封装成一个函数 isHasObject
            if(!_this.isHasObject(prev, key)) {
                prev.push(key);
            }
            return prev
        }, [])
    }
}

getObject是将name转换为对应的对象{key:value},拿到key之后,判断数组里是否已经含有了key,有的话就跳过,否则把key放入到数组里。

class GroupMultiple {
    //data是聚合的数据, name是聚合的字段
    constructor (data, name){
        this.data = data;
        this.name = name;
    }

    //获取key  
    getKeys(data, names){
        let _this = this;
        return data.reduce(function (prev, cur) {
            //首先拿到name对应的值, 将key组成 {key: value}形式
            var key = _this.getObject(cur, names);
            //判断数组里是否已经含有了key,有的话就跳过,否则把key放入到数组里
            //把判断方法封装成一个函数 isHasObject
            if(!_this.isHasObject(prev, key)) {
                prev.push(key);
            }
            return prev
        }, [])
    }

    getObject(obj, keys=this.name) {
        let aa = {};
        for (var i = 0, length = keys.length; i < length; i++) {
            aa[keys[i]] = obj[keys[i]]
        }
        return aa
    }

    //判断数组中是否有了某个对象
    isHasObject(prev, keys){
        //数组长度为0的时候直接返回
        if (prev.length === 0) return false;
        var _this = this;
        //判断数组中的每一项是否与keys相等, 这里用到数组的find方法,当数组的每一项与keys都不想等的时候(也就是数组中没有这个对象)返回undefined,
        //isObjectValueEqual 是判断两个对象是否相等的方法
        return prev.find(function (item) {
            return _this.isObjectValueEqual(item, keys)
        })
    }

    //判断两个对象是否相等
    isObjectValueEqual(obj1, obj2) {
        // 只有两个对象的key和value完全相等时这两个对象才相等
        var _this = this;
        var keys1 = Object.keys(obj1);
        var keys2 = Object.keys(obj2);
        if (keys1.length !== keys2.length) {
            return false;
        }
        for (var i = 0; i < keys1.length; i++) {
            var key = keys1[i];
            if (obj1[key] !== obj2[key]) {
                return false;
            }
        }
        return true;
    }
}

拿到对应的key之后,需要把每一个key在数组中对应的值取出来

class GroupMultiple {
    ...
    //返回一个数组中包含某个对象的元素
    hasObjectData(data, obj){
        return data.filter(function (item) {
            return Object.keys(obj).reduce(function(pr, cr){
                if (pr){
                    return obj[cr] === item[cr]
                } else {
                    return false
                }
            }, true)
        });
    }
}

然后将结果返回

class GroupMultiple {
    ...
    //返回一个数组中包含某个对象的元素
    hasObjectData(data, obj){
        return data.filter(function (item) {
            return Object.keys(obj).reduce(function(pr, cr){
                if (pr){
                    return obj[cr] === item[cr]
                } else {
                    return false
                }
            }, true)
        });
    }

    //返回结果
    init(){
        let _this = this;
        //拿到keys
        let keys = _this.getKeys(_this.data, this.name);
        return keys.map(function(item) {
            //拿到keys对应的值
            let vals = _this.hasObjectData(_this.data, item);
            return {
                key: item,
                value: vals
            };
        });
    }
}
var group_result = new GroupMultiple (arr, ['age', 'score']);
console.log(group_result .init())

[{…}, {…}, {…}, {…}]
    0:
        key: {age: 18, score: 60}
        value: Array(3)
            0: {name: "小孙", age: 18, score: 60, weight: 60}
            1: {name: "小李", age: 18, score: 60, weight: 70}
            2: {name: "小赵", age: 18, score: 60, weight: 60}
    1:
        key: {age: 19, score: 70}
        value: Array(2)
            0: {name: "小王", age: 19, score: 70, weight: 55}
            1: {name: "小钱", age: 19, score: 70, weight: 55}
    2:
        key: {age: 20, score: 70}
        value: Array(1)
            0: {name: "小刘", age: 20, score: 70, weight: 65}
    
    3:
        key: {age: 20, score: 60}
        value: Array(1)
            0: {name: "小周", age: 20, score: 60, weight: 50}

到这里多个字段的聚合算是完成了。但是有个问题,要是我想聚合之后做其他的操作呢?比如说求和,求平均等。

group之后对数据的再处理

例如求和:
增加一个sum方法,并且在增加要操作的对象operating,修改init方法

class GroupMultiple {
    constructor (data, name, operating=null){
        this.data = data;
        this.name = name;
        this.operating = operating;
    }
    ...
    //返回结果
    init(){
        let _this = this;
        //拿到keys
        let keys = _this.getKeys(_this.data, this.name);
        return keys.map(function(item) {
            //拿到keys对应的值
            let vals = _this.hasObjectData(_this.data, item);
            if (_this.operating){
                Object.keys(_this.operating).forEach(opitem => {
                    let ret = _this[_this.operating[opitem]](opitem, vals);
                    vals.forEach(val => {
                        val[opitem] = ret
                    })
                })
            }
            return {
                key: item,
                value: vals
            };
        });
    }
    sum(key, data){
        //key是要求和的字段, data是要求和的数组
        return data.reduce(function (prev, cur) {
            return prev + cur[key]
        }, 0)
    }
}
//对 age 和 score聚合, 然后对weight求和
var sum_group = new GroupMultiple(arr, ['age', 'score'], {weight: 'sum'});
console.log(sum_group.init())
[{…}, {…}, {…}, {…}]
    0:
        key: {age: 18, score: 60}
        value: Array(3)
            0: {name: "小孙", age: 18, score: 60, weight: 190}
            1: {name: "小李", age: 18, score: 60, weight: 190}
            2: {name: "小赵", age: 18, score: 60, weight: 190}
    1:
        key: {age: 19, score: 70}
        value: Array(2)
            0: {name: "小王", age: 19, score: 70, weight: 110}
            1: {name: "小钱", age: 19, score: 70, weight: 110}
    2:
        key: {age: 20, score: 70}
        value: Array(1)
            0: {name: "小刘", age: 20, score: 70, weight: 65}
    
    3:
        key: {age: 20, score: 60}
        value: Array(1)
            0: {name: "小周", age: 20, score: 60, weight: 50}

可以看到每一项的weight都成为该分组中weight的和了。
同理还可以增加 mean(),max(),min(),first(),end()等。不过不建议这么做,处理数据最好还是在后台处理,js毕竟对数据处理的支持不到位。
源代码: https://github.com/controZheng/work

你可能感兴趣的:(js使用groupby)