Question

我有一个问题：

function foo() {
  this.a = 1;

  this.bar = function () {
    return this.a;
  }
}
var instance = new foo;
instance.bar(); // returns 1
instance.bar.bind(otherContext)(); // returns 2

我需要的是输出每一行，但我无法弄清楚如何做到这一点。现在，我得到了这个：

SELECT name, APN, BPN, count(APN), min(aCost), min(bCost), ceil(avg(aQty)), 
max(aQty), sum(bShipped), 
CONCAT(truncate((avg(aResale)-avg(aCost))/avg(aResale),2) * 100,'%'), code FROM
(SELECT name, APN, BPN, aCost, aQty, code
    FROM table_1
    WHERE customer = '12345' AND
    aDate >= '2013-01-01' and
    aDate <= '2015-12-12') as qh
INNER JOIN (SELECT CPN, bCost, bResale, bShipped from table_2 
WHERE customer = '12345') as ih
ON qh.APN = ih.CPN
WHERE bShipped > 0
GROUP BY qh.APN;

由于NAME | APN | BPN | APN Count asdf 001 555 3 /*summary of APN 001*/ qere 002 865 1 /*summary of APN 002*/ rtrt 003 123 2 /*summary of APN 003*/，我获得group by的摘要，但是如果我不使用任何qh.APN语句，我的结果只有1行。

我希望有这个group by列，而同时显示所有行 - 没有汇总任何值。像这样：

APN Count

我需要查看构成NAME | APN | BPN | APN Count asdf 001 555 3 /*Individual record, with count too*/ asdf 001 862 3 /*Individual record, with count too*/ asdf 001 999 3 /*Individual record, with count too*/ qere 002 865 1 /*Individual record, with count too*/ rtrt 003 123 2 /*Individual record, with count too*/ rtrt 003 456 2 /*Individual record, with count too*/列的每条记录，因为我需要查看每个APN Count，而不只是一个摘要行。我现在用BPN写这个，因为看到汇总数据比 nothing 更好，因为我不知道在这里使用正确的语法来得到结果我想。

Answer 1

将表1与您的聚合查询作为子查询连接。

SELECT t1.name, t1.apn, t1.bpn, t1.code, t2.*
FROM table_1 AS t1
JOIN (
    SELECT APN, count(APN) AS APN_count, min(aCost) AS min_aCost, min(bCost) AS min_bCost, ceil(avg(aQty)) AS avgQty, 
    max(aQty) AS maxQty, sum(bShipped) AS sum_bShipped, 
    CONCAT(truncate((avg(aResale)-avg(aCost))/avg(aResale),2) * 100,'%') AS avg_Margin FROM
    (SELECT name, APN, BPN, aCost, aQty, code
        FROM table_1
        WHERE customer = '12345' AND
        aDate >= '2013-01-01' and
        aDate <= '2015-12-12') as qh
    INNER JOIN (SELECT CPN, bCost, bResale, bShipped from table_2 
    WHERE customer = '12345') as ih
    ON qh.APN = ih.CPN
    WHERE bShipped > 0
    GROUP BY qh.APN) AS t2
ON t1.APN = t2.APN

Answer 2

显而易见的方法是将结果集加入某些原始表。在你的情况下，我不确定什么是原始的，但它可能是这样的：

SELECT t.*,
       t_agregated.*
FROM table_1 t
LEFT JOIN (
SELECT t1.APN,  
       count(t1.APN), 
       min(t1.aCost), 
       min(bCost), 
       ceil(avg(t1.aQty)), 
       max(aQty), 
       sum(bShipped), 
       CONCAT(truncate((avg(t1.aResale)-avg(t1.aCost))/avg(t1.aResale),2) * 100,'%')
FROM table_1 t1
INNER JOIN table_2 t2
ON t1.APN = t2.CPN
   AND t2.bShipped > 0
WHERE t1.customer = '12345' AND
    aDate >= '2013-01-01' and
    aDate <= '2015-12-12'
GROUP BY t1.APN) t_agregated
ON t.APN = t_agregated.APN
WHERE t.customer = '12345' AND
    t.aDate >= '2013-01-01' and
    t.aDate <= '2015-12-12'

但我很确定，你的目标远非如此。您应该更好地发布原始数据源和预期结果集。在这里提问的好方法是为sqlfiddle提供表格方案和数据样本。

mysql同时使用摘要检索所有行

2 个答案: