（13）mongodb 聚集运算之 group

mongodb有3个方法实现数据的分析与统计，分别是group分组统计、aggregate简单聚合、mapReduce强大统计。group 函数需要自己实现逻辑，所以它很灵活，但是有一个缺点就是不支持分片，不支持数据的分布式运算，在下面的章节中再分析支持share分片的方法。下面先说一下 group 函数的语法构成：

db.collection.group(document)

document格式：
{
key:{key1:1,key2:1},
cond:{},
reduce:function(curr,result){
},
initial:{},
finalize:function(){
}
}

说明：

key：分组字段，是对谁进行分组，相当于关系型数据库中的group by。

cond：查询条件，相当于where。

reduce：聚合函数，是自己的实现逻辑，也就是在分组的结果中做什么。

initial：初始化，可以理解为进入分组先做什么。

finalize：统计一组后的回调函数，理解为跳出分组后做什么，在reduce后执行。

下面用具体列子说明。

1、准备数据：将下面31条数据写入shop库的goods表

[{"goods_id":1,"cat_id":4,"goods_name":"KD876","goods_number":1,"click_count":7,"shop_price":1388.00,"add_time":1240902890},{"goods_id":4,"cat_id":8,"goods_name":"u8bfau57fau4e9aN85u539fu88c5u5145u7535u5668","goods_number":17,"click_count":0,"shop_price":58.00,"add_time":1241422402},{"goods_id":3,"cat_id":8,"goods_name":"u8bfau57fau4e9au539fu88c55800u8033u673a","goods_number":24,"click_count":3,"shop_price":68.00,"add_time":1241422082},{"goods_id":5,"cat_id":11,"goods_name":"u7d22u7231u539fu88c5M2u5361u8bfbu5361u5668","goods_number":8,"click_count":3,"shop_price":20.00,"add_time":1241422518},{"goods_id":6,"cat_id":11,"goods_name":"u80dcu521bKINGMAXu5185u5b58u5361","goods_number":15,"click_count":0,"shop_price":42.00,"add_time":1241422573},{"goods_id":7,"cat_id":8,"goods_name":"u8bfau57fau4e9aN85u539fu88c5u7acbu4f53u58f0u8033u673aHS-82","goods_number":20,"click_count":0,"shop_price":100.00,"add_time":1241422785},{"goods_id":8,"cat_id":3,"goods_name":"u98deu5229u6d669@9v","goods_number":1,"click_count":9,"shop_price":399.00,"add_time":1241425512},{"goods_id":9,"cat_id":3,"goods_name":"u8bfau57fau4e9aE66","goods_number":4,"click_count":20,"shop_price":2298.00,"add_time":1241511871},{"goods_id":10,"cat_id":3,"goods_name":"u7d22u7231C702c","goods_number":7,"click_count":11,"shop_price":1328.00,"add_time":1241965622},{"goods_id":11,"cat_id":3,"goods_name":"u7d22u7231C702c","goods_number":1,"click_count":0,"shop_price":1300.00,"add_time":1241966951},{"goods_id":12,"cat_id":3,"goods_name":"u6469u6258u7f57u62c9A810","goods_number":8,"click_count":13,"shop_price":983.00,"add_time":1245297652}]

[{"goods_id":13,"cat_id":3,"goods_name":"u8bfau57fau4e9a5320 XpressMusic","goods_number":8,"click_count":13,"shop_price":1311.00,"add_time":1241967762},{"goods_id":14,"cat_id":4,"goods_name":"u8bfau57fau4e9a5800XM","goods_number":1,"click_count":6,"shop_price":2625.00,"add_time":1241968492},{"goods_id":15,"cat_id":3,"goods_name":"u6469u6258u7f57u62c9A810","goods_number":3,"click_count":8,"shop_price":788.00,"add_time":1241968703},{"goods_id":16,"cat_id":2,"goods_name":"u6052u57fau4f1fu4e1aG101","goods_number":0,"click_count":3,"shop_price":823.33,"add_time":1241968949},{"goods_id":17,"cat_id":3,"goods_name":"u590fu65b0N7","goods_number":1,"click_count":2,"shop_price":2300.00,"add_time":1241969394},{"goods_id":18,"cat_id":4,"goods_name":"u590fu65b0T5","goods_number":1,"click_count":0,"shop_price":2878.00,"add_time":1241969533},{"goods_id":19,"cat_id":3,"goods_name":"u4e09u661fSGH-F258","goods_number":12,"click_count":7,"shop_price":858.00,"add_time":1241970139},{"goods_id":20,"cat_id":3,"goods_name":"u4e09u661fBC01","goods_number":12,"click_count":14,"shop_price":280.00,"add_time":1241970417},{"goods_id":21,"cat_id":3,"goods_name":"u91d1u7acb A30","goods_number":40,"click_count":4,"shop_price":2000.00,"add_time":1241970634},{"goods_id":22,"cat_id":3,"goods_name":"u591au666eu8fbeTouch HD","goods_number":1,"click_count":15,"shop_price":5999.00,"add_time":1241971076}]

[{"goods_id":23,"cat_id":5,"goods_name":"u8bfau57fau4e9aN96","goods_number":8,"click_count":17,"shop_price":3700.00,"add_time":1241971488},{"goods_id":24,"cat_id":3,"goods_name":"P806","goods_number":100,"click_count":35,"shop_price":2000.00,"add_time":1241971981},{"goods_id":25,"cat_id":13,"goods_name":"u5c0fu7075u901a/u56fau8bdd50u5143u5145u503cu5361","goods_number":2,"click_count":0,"shop_price":48.00,"add_time":1241972709},{"goods_id":26,"cat_id":13,"goods_name":"u5c0fu7075u901a/u56fau8bdd20u5143u5145u503cu5361","goods_number":2,"click_count":0,"shop_price":19.00,"add_time":1241972789},{"goods_id":27,"cat_id":15,"goods_name":"u8054u901a100u5143u5145u503cu5361","goods_number":2,"click_count":0,"shop_price":95.00,"add_time":1241972894},{"goods_id":28,"cat_id":15,"goods_name":"u8054u901a50u5143u5145u503cu5361","goods_number":0,"click_count":0,"shop_price":45.00,"add_time":1241972976},{"goods_id":29,"cat_id":14,"goods_name":"u79fbu52a8100u5143u5145u503cu5361","goods_number":0,"click_count":0,"shop_price":90.00,"add_time":1241973022},{"goods_id":30,"cat_id":14,"goods_name":"u79fbu52a820u5143u5145u503cu5361","goods_number":9,"click_count":1,"shop_price":18.00,"add_time":1241973114},{"goods_id":31,"cat_id":3,"goods_name":"u6469u6258u7f57u62c9E8 ","goods_number":1,"click_count":5,"shop_price":1337.00,"add_time":1242110412},{"goods_id":32,"cat_id":3,"goods_name":"u8bfau57fau4e9aN85","goods_number":4,"click_count":9,"shop_price":3010.00,"add_time":1242110760}]

2、示例：查询每个栏目下的商品数据

> db.goods.group({
... key:{cat_id:1},
... cond:{},
... reduce:function(curr,result){
... result.cnt+=1;
... },
... initial:{cnt:0}
... })

结果：

[
    {
        "cat_id" : 4,
        "cnt" : 3
    },
    {
        "cat_id" : 8,
        "cnt" : 3
    },
    {
        "cat_id" : 11,
        "cnt" : 2
    },
    {
        "cat_id" : 3,
        "cnt" : 15
    },
    {
        "cat_id" : 2,
        "cnt" : 1
    },
    {
        "cat_id" : 5,
        "cnt" : 1
    },
    {
        "cat_id" : 13,
        "cnt" : 2
    },
    {
        "cat_id" : 15,
        "cnt" : 2
    },
    {
        "cat_id" : 14,
        "cnt" : 2
    }
]

3、示例：查询每个栏目下价格高于50元的商品数量

> db.goods.group({
... key:{cat_id:1},
... cond:{shop_price:{$gt:50}},
... reduce:function(curr,result){
... result.cnt+=1;
... },
... initial:{cnt:0}
... })

结果：

[
    {
        "cat_id" : 4,
        "cnt" : 3
    },
    {
        "cat_id" : 8,
        "cnt" : 3
    },
    {
        "cat_id" : 3,
        "cnt" : 15
    },
    {
        "cat_id" : 2,
        "cnt" : 1
    },
    {
        "cat_id" : 5,
        "cnt" : 1
    },
    {
        "cat_id" : 15,
        "cnt" : 1
    },
    {
        "cat_id" : 14,
        "cnt" : 1
    }
]

4、示例：查询每个栏目下的商品库存量

> db.goods.group({
... key:{cat_id:1},
... cond:{},
... reduce:function(curr,result){
... result.num+=curr.goods_number
... },
... initial:{
... num:0
... }
... }
... )

结果：

[
    {
        "cat_id" : 4,
        "num" : 3
    },
    {
        "cat_id" : 8,
        "num" : 61
    },
    {
        "cat_id" : 11,
        "num" : 23
    },
    {
        "cat_id" : 3,
        "num" : 203
    },
    {
        "cat_id" : 2,
        "num" : 0
    },
    {
        "cat_id" : 5,
        "num" : 8
    },
    {
        "cat_id" : 13,
        "num" : 4
    },
    {
        "cat_id" : 15,
        "num" : 2
    },
    {
        "cat_id" : 14,
        "num" : 9
    }
]

5、示例：查询每个栏目下最贵的商品

> db.goods.group({
... key:{cat_id:1},
... cond:{},
... reduce:function(curr,result){
...   if(result.max<curr.shop_price){
...      result.max=curr.shop_price;
...   }
... },
... initial:{max:0}
... })

结果：

[
    {
        "cat_id" : 4,
        "max" : 2878
    },
    {
        "cat_id" : 8,
        "max" : 100
    },
    {
        "cat_id" : 11,
        "max" : 42
    },
    {
        "cat_id" : 3,
        "max" : 5999
    },
    {
        "cat_id" : 2,
        "max" : 823.33
    },
    {
        "cat_id" : 5,
        "max" : 3700
    },
    {
        "cat_id" : 13,
        "max" : 48
    },
    {
        "cat_id" : 15,
        "max" : 95
    },
    {
        "cat_id" : 14,
        "max" : 90
    }
]

6、示例：查询每个栏目下商品的平均价格

> db.goods.group({
...  key:{cat_id:1},
...  cond:{},
...  reduce:function(curr,result){
...   result.cnt+=1;
...   result.sum+=curr.shop_price;
... },
... initial:{sum:0,cnt:0},
... finalize:function(result){
... result.avg=result.sum/result.cnt;
... }
... })

结果：

[
    {
        "cat_id" : 4,
        "sum" : 6891,
        "cnt" : 3,
        "avg" : 2297
    },
    {
        "cat_id" : 8,
        "sum" : 226,
        "cnt" : 3,
        "avg" : 75.33333333333333
    },
    {
        "cat_id" : 11,
        "sum" : 62,
        "cnt" : 2,
        "avg" : 31
    },
    {
        "cat_id" : 3,
        "sum" : 26191,
        "cnt" : 15,
        "avg" : 1746.0666666666666
    },
    {
        "cat_id" : 2,
        "sum" : 823.33,
        "cnt" : 1,
        "avg" : 823.33
    },
    {
        "cat_id" : 5,
        "sum" : 3700,
        "cnt" : 1,
        "avg" : 3700
    },
    {
        "cat_id" : 13,
        "sum" : 67,
        "cnt" : 2,
        "avg" : 33.5
    },
    {
        "cat_id" : 15,
        "sum" : 140,
        "cnt" : 2,
        "avg" : 70
    },
    {
        "cat_id" : 14,
        "sum" : 108,
        "cnt" : 2,
        "avg" : 54
    }
]

 　　可以看出最后返回的json格式的键包括了分组的字段和result中用的字段。在initial中初始化的键，reduce中可以用，结果中可以返回，聚合函数reduce有两个参数，一个是curr，可以根据当前遍历数据的键取值，一个是result，可以取初始的值进行运算，不断的修改这个值。回调函数finalize有一个参数result，接收reduce计算的结果。