内部联接非常慢。我该如何优化？

Question

我想在时间属性上找到多个查询的交集。

我尝试了所有这些对象的内部连接，但是这花费了大量的时间，而且从未完成。

    SELECT 
      pdx.time, 
      demand,
      pdx.temperature, 
      sea.temperature, 
      boi.temperature, 
      geg.temperature
    FROM capstone.aggregate_power_demands
    INNER JOIN  
      ( 
        SELECT time, temperature
        FROM capstone.weather
        WHERE location_name='PDX' AND time >= '1993' AND time <= '2018'
      ) as pdx
    ON aggregate_power_demands.time = pdx.time 
    INNER JOIN 
      ( 
        SELECT time, temperature 
        FROM capstone.weather
        WHERE location_name='SEA' AND time >= '1993' AND time <= '2018'
      ) as sea
    ON pdx.time = sea.time 
    INNER JOIN 
      ( 
        SELECT time, temperature 
        FROM capstone.weather
        WHERE location_name='BOI' AND time >= '1993' AND time <= '2018'
      ) as boi
    ON sea.time = boi.time
    INNER JOIN 
      ( 
        SELECT time, temperature
        FROM capstone.weather
        WHERE location_name='GEG' AND time >= '1993' AND time <= '2018'
      ) as geg
    ON boi.time = geg.time

Answer 1

您也可以尝试一下。它只有一个连接：

SELECT 
      agd.time
    , agd.demand
    , MAX( IF( w.location_name = 'PDX', w.temperature, NULL)) as pdx_temperature
    , MAX( IF( w.location_name = 'SEA', w.temperature, NULL)) as sea_temperature
    , MAX( IF( w.location_name = 'BOI', w.temperature, NULL)) as boi_temperature
    , MAX( IF( w.location_name = 'GEG', w.temperature, NULL)) as geg_temperature
FROM capstone.aggregate_power_demands agd
LEFT JOIN capstone.weather w ON agd.time = w.time
WHERE w.location_name IN ( 'PDX' , 'SEA' , 'BOI' , 'GEG' )
AND agd.time BETWEEN '1993-01-01 00:00:00' AND '2018-12-31 23:59:59'
GROUP BY w.location_name;

Answer 2

尝试一下：

SELECT 
  aggregate_power_demands.time, 
  demand,
  pdx.temperature, 
  sea.temperature, 
  boi.temperature, 
  geg.temperature
FROM capstone.aggregate_power_demands
LEFT JOIN  
  (SELECT time, temperature
    FROM capstone.weather
    WHERE location_name='PDX' AND YEAR(time) BETWEEN '1993' AND '2018' 
GROUP BY location_name,time) as pdx
ON aggregate_power_demands.time = pdx.time 
LEFT JOIN 
  (SELECT time, temperature 
    FROM capstone.weather
    WHERE location_name='SEA' AND YEAR(time) BETWEEN '1993' AND '2018' 
GROUP BY location_name,time) as sea
ON aggregate_power_demands.time = sea.time 
LEFT JOIN 
  (SELECT time, temperature 
    FROM capstone.weather
    WHERE location_name='BOI' AND YEAR(time) BETWEEN '1993' AND '2018' 
GROUP BY location_name,time) as boi
ON aggregate_power_demands.time = boi.time
LEFT JOIN 
  (SELECT time, temperature
    FROM capstone.weather
    WHERE location_name='GEG' AND YEAR(time) BETWEEN '1993' AND '2018' 
GROUP BY location_name,time) as geg
ON aggregate_power_demands.time = geg.time;