如何让我的查询更快地选择变量答案

【问题标题】：how to make my query select variables faster如何让我的查询更快地选择变量
【发布时间】：2021-06-15 19:04:27
【问题描述】：

我有 sql 查询，必须快速运行

查询正在工作，但速度很慢我怎样才能让它们工作得更快

这是我的第一个查询，对于我尝试做的事情来说非常慢

select 
distinct symbol as s
,(select count(distinct login) from orders where symbol = s and cmd < 2)as users
,(select count(cmd) from orders where symbol=s and cmd=1) as sat
,(select count(cmd) from orders where symbol=s and cmd=0) as al
,(select ifnull(sum(volume),0) from orders where symbol=s and cmd=0) as al_lot
,(select ifnull(sum(volume),0) from orders where symbol=s and cmd=1) as sat_lot
,(select count(distinct login) from orders where symbol=s and cmd=0) as a
,(select count(distinct login) from orders dd where symbol=s and (select sum(volume) from orders where symbol=dd.symbol and orders.login=dd.login and cmd = 0) < (select sum(volume) from orders where symbol=dd.symbol and dd.login=orders.login and cmd=1)) as b
from orders where symbol in("EURUSD","XAUUSD","GAUTRY","GBPUSD","USDJPY","USDTRY","BRENT","CRUDE","DAX30","DM#","ES#","BTCUSD") group by s,cmd;

这是我目前正在使用的查询（我试图让它更快我的第一个查询）

Select 
src.symbol as Symbol
,IFNULL(count(distinct deals.LOGIN),0) as Users
,IFNULL(count(mybuy.LOGIN),0) as BuyDeals
,IFNULL(count(mysell.LOGIN),0) as SellDeals
,IFNULL(sum(mybuy.volume),0) as BuyVolume
,IFNULL(sum(mysell.volume),0) as SellVolume
,(select count(distinct login) from orders where symbol=deals.symbol and cmd =0) as a
,(select count(distinct login) from orders dd where symbol=src.symbol and (select sum(volume) from orders where symbol=dd.symbol and login=dd.login and cmd = 0) < (select sum(volume) from orders where symbol=dd.symbol and dd.login=login and cmd=1)) as b

from
(Select * from orders Where cmd<2)src
LEFT OUTER JOIN(Select `order`, symbol, login, volume from orders where cmd<2 group by `order`,symbol,login)deals on deals.`order`=src.`order` 
LEFT OUTER JOIN(Select `order`, symbol, login, volume from orders where cmd=0 group by `order`,symbol,login)mybuy on mybuy.`order`=src.`order`
LEFT OUTER JOIN(Select `order`, symbol, login, volume from orders where cmd=1 group by `order`,symbol,login)mysell on mysell.`order`=src.`order`

where src.symbol in ('EURUSD','XAUUSD','GAUTRY','GBPUSD','USDJPY','USDTRY','BRENT','CRUDE','DAX30','DM#','ES#','BTCUSD') group by src.symbol order by symbol;

但它仍然太慢，它必须更快地工作

这是我的桌子

CREATE TABLE `orders` (
  `order` int(11) NOT NULL,
  `login` int(11) DEFAULT NULL,
  `symbol` varchar(13) DEFAULT NULL,
  `digits` int(11) DEFAULT NULL,
  `cmd` int(11) DEFAULT NULL,
  `volume` int(11) DEFAULT NULL,
  `open__time` varchar(45) DEFAULT NULL,
  `state` int(11) DEFAULT NULL,
  `open_price` double DEFAULT NULL,
  `sl` double DEFAULT NULL,
  `tp` double DEFAULT NULL,
  `close_time` varchar(45) DEFAULT NULL,
  `gw_volume` int(11) DEFAULT NULL,
  `expiration` varchar(45) DEFAULT NULL,
  `reason` varchar(45) DEFAULT NULL,
  `conv_reserv` varchar(3) DEFAULT NULL,
  `conv_rate1` double DEFAULT NULL,
  `conv_rate2` double DEFAULT NULL,
  `commission` double DEFAULT NULL,
  `commission_agent` double DEFAULT NULL,
  `storage` double DEFAULT NULL,
  `close_price` double DEFAULT NULL,
  `profit` double DEFAULT NULL,
  `taxes` double DEFAULT NULL,
  `magic` int(11) DEFAULT NULL,
  `comment` varchar(45) DEFAULT NULL,
  `gw_order` int(11) DEFAULT NULL,
  `activation` int(11) DEFAULT NULL,
  `gw_open_price` int(11) DEFAULT NULL,
  `gw_close_price` int(11) DEFAULT NULL,
  `margin_rate` double DEFAULT NULL,
  `timestamp` varchar(45) DEFAULT NULL,
  PRIMARY KEY (`order`)
) ENGINE=InnoDB DEFAULT CHARSET=utf8;

【问题讨论】：

您必须删除所有子查询。
但我需要他们给我的价值观@Akina
您必须将它们转换为 JOIN 形式。
第二个我删除了其中一些，但我不知道如何在我的第二个查询子查询@Akina 中删除和获取变量
我尝试使用 LEFT OUTER JOIN 获取数据，所以我不需要使用子查询 @Akina

标签： mysql sql database performance time

【解决方案1】：

使用条件聚合！

select symbol,
       count(distinct case when cmd < 2 then login) as users,
       sum(cmd = 1) as sat,
       sum(cmd = 0) as al,
       sum(case when cmd = 0 then volume else 0 end) as al_lot,
       sum(case when cmd = 1 then volume else 0 end) as sat_lot,
       count(distinct case when cmd = 0 then login) as a
from orders
where symbol in ('EURUSD', 'XAUUSD', 'GAUTRY', 'GBPUSD', 'USDJPY', 'USDTRY',
                 'BRENT', 'CRUDE', 'DAX30', 'DM#', 'ES#', 'BTCUSD'
                )
group by symbol;

这不包括引用dd 的最后一列。我在任何地方都没有看到dd 的定义，所以看起来它会导致错误。在任何情况下，您都应该能够使用具有聚合函数的表达式来构造该逻辑。

【讨论】：