Question

我有一张包含四个不同数字字段的行表：人物，碎片，英里和行李。所有这些字段现在每行都为0，因为还没有数据。这些行还有两个名为“County”的重要字段，其值是纽约州的县之一，而“Underwater”则只是1/0（是/否）字段。所以他们看起来像这样...

County      Underwater  People Debris Miles Bags
Richmond             0       0      0     0    0
Rockland             1       0      0     0    0
Nassau               0       0      0     0    0
Queens               1       0      0     0    0
Niagara              0       0      0     0    0

为了将所有这些组合在一起用于PHP页面，我使用了一个包含大量UNION的长SQL语句，并且基本上将结果分别来自纽约市的县，来自长岛的县和Upstate New约克，然后另外两个水下而不是水下的查询，然后是一个总计。我认为使用一次的长语句比对同一个表的六个单独查询更有效，它会省去一些代码行。我在所有数据为0之前使用的是测试数字，并且工作正常。现在所有数据都是0，结果集是一行全0，然后是一行NULL，没有别的。因为我使用foreach PHP循环解析数据，所以它与我的算法紧密相关。我还注意到，如果我将1行的数据（即使只有1个字段）更改为非零，那么我将在结果集中获得另一行。

任何人都可以解释这种行为和解决方案或者提供替代方案吗？如果需要，我可以更深入。

顺便说一句，这是有问题的陈述。待命墙......

SELECT SUM( people ) AS people, SUM( debris ) AS debris, SUM( miles ) AS miles, SUM( bags ) AS bags 
FROM `table` 
WHERE county='The Bronx' OR county='Kings' OR county='New York' OR county='Queens' OR county='Richmond' 
UNION 
SELECT SUM( people ) AS people, SUM( debris ) AS debris, SUM( miles ) AS miles, SUM( bags ) AS bags 
FROM `table` 
WHERE county='Nassau' OR county='Suffolk' 
UNION 
SELECT SUM( people ) AS people, SUM( debris ) AS debris, SUM( miles ) AS miles, SUM( bags ) AS bags 
FROM `table` 
WHERE county='Albany' OR county='Allegany' OR county='Broome' OR county='Cattaraugus' OR county='Cayuga' OR county='Chautauqua' OR county='Chemung' OR county='Chenango' OR county='Clinton' OR county='Columbia' OR county='Cortland' OR county='Delaware' OR county='Dutchess' OR county='Erie' OR county='Essex' OR county='Franklin' OR county='Fulton' OR county='Genesee' OR county='Greene' OR county='Hamilton' OR county='Herkimer' OR county='Jefferson' OR county='Lewis' OR county='Livingston' OR county='Madison' OR county='Monroe' OR county='Montgomery' OR county='Niagara' OR county='Oneida' OR county='Onondaga' OR county='Ontario' OR county='Orange' OR county='Orleans' OR county='Oswego' OR county='Otsego' OR county='Putnam' OR county='Rensselaer' OR county='Rockland' OR county='Saratoga' OR county='Schenectady' OR county='Schoharie' OR county='Schuyler' OR county='Seneca' OR county='Steuben' OR county='StLawrence' OR county='Sullivan' OR county='Tioga' OR county='Tompkins' OR county='Ulster' OR county='Warren' OR county='Washington' OR county='Wayne' OR county='Westchester' OR county='Wyoming' OR county='Yates' 
UNION 
SELECT SUM( people ) AS people, SUM( debris ) AS debris, SUM( miles ) AS miles, SUM( bags ) AS bags 
FROM `table` 
WHERE underwater = '0' AND county != '' 
UNION 
SELECT SUM( people ) AS people, SUM( debris ) AS debris, SUM( miles ) AS miles, SUM( bags ) AS bags 
FROM `table` 
WHERE underwater = '1' AND county != '' 
UNION 
SELECT SUM( people ) AS people, SUM( debris ) AS debris, SUM( miles ) AS miles, SUM( bags ) AS bags 
FROM `table` 
WHERE county != '' AND fname != '' AND lname != ''

Answer 1

使用UNION ALL代替UNION。 UNION运算符会在两个结果集之间抛出重复的行，就像SELECT DISTINCT对单个结果集一样。（或者也选择县名，因为这会导致结果行包含不相同的数据。）

要从查询中消除NULL行（这是由空数据集上的SUM()等聚合函数的结果），只需执行以下操作：

SELECT * FROM (
    your query here
) AS x
WHERE x.people IS NOT NULL

或者，将查询中SUM(x)的每次出现更改为COALESCE(SUM(x), 0)。

UNION中带有6个SELECT的SQL语句仅返回2行

1 个答案: