1、当我们要统计数据表数量时，如果遇到多表查询，会出现一个主表对应多个子表的维度，我们会用到group by，但是不要再用统计函数去操作数据，因为统计还是会统计原数据
案例
SELECT
sum(`order`.goods_amount) AS total_goods_amount,
sum(`order`.order_amount) AS total_order_amount,
sum(`order`.redbag_amount) AS total_redbag_amount,
sum(`order`.spread_amount) AS total_spread_amount,
sum(`order`.redrain_amount) AS total_redrain_amount
FROM
ds_order AS `order`
LEFT JOIN ds_refundreturn AS refundreturn ON `order`.order_id = refundreturn.order_id
WHERE
1 = 1
AND `order`.order_state = '40'
AND refundreturn.refund_state < '3'
AND refundreturn.is_revoke = '0'
GROUP BY
`order`.order_id
这样得到的数据，会把子表多个订单ID给统计上去，而分组并没有限制到
2、可以用数据库子查询，最后再加一个总查询
select sum(a.total_goods_marketprice) from (
SELECT
(
goods.goods_marketprice * ordergoods.goods_num
) AS total_goods_marketprice
FROM
ds_order AS `order`
LEFT JOIN ds_refundreturn AS refundreturn ON `order`.order_id = refundreturn.order_id
LEFT JOIN ds_ordergoods AS ordergoods ON `order`.order_id = ordergoods.order_id
LEFT JOIN ds_goods AS goods ON ordergoods.goods_id = goods.goods_id
WHERE
1 = 1
AND `order`.order_state = '40'
AND refundreturn.refund_state < '3'
AND refundreturn.is_revoke = '0'
GROUP BY
`order`.order_id
)a;
3、mysql group concat 长度是有限制的，如果数据太多，会截取
4、如果在不同电脑查询值，排序值得到的数据是相同的，有可能会显示不同顺序的数据
5、mysql看or与in的条件是一样的，但是数量大的时候，or会影响性能
亲，后面还会陆续更新！