Question

我有两张桌子。一个商店＆＃34;地点＆＃34;：

TABLE location (
    ID               serial             PRIMARY KEY,
    name             text               NOT NULL,
    description      text               NOT NULL
);

每个位置都有多行＆＃34;数据＆＃34;：

TABLE data(
    ID               smallint           REFERENCES location(ID),
    date             date,               
    rainfall         int                
);

我想找到所有包含＆＃34;数据＆＃34;跨越一定时期，至少有＃34; min＆＃34;该期间内的值数量。我试过这个：

SELECT location.ID, location.name 
FROM location 
JOIN data 
ON data.id = location.id 
GROUP BY location.id 
HAVING MIN(data.date) <= '$start_date' 
       AND
       MAX(data.date) >= '$end_date'  
       AND 
       (SELECT COUNT(*) FROM data WHERE data.date >= '$start_date' AND data.date <= '$end_date') >= '$min' 
ORDER BY location.ID

如果我取出第二个最后一行（SELECT COUNT行），它会正确返回数据跨越所需期间的位置（但不具有所需期间内的值数量大于约束的约束）或等于＆＃34; min＆＃34;）。

谁能告诉我如何施加约束？即。我的＆＃34; SELECT COUNT行＆＃34;。

出了什么问题

以下示例数据可能有助于澄清我的问题：

示例数据：

location:
ID = 1, name = "London",   description = "test location 1"
ID = 2, name = "New York", description = "test location 2"

数据：

ID = 1, date = 2001-01-01, rainfall = 0.0
ID = 1, date = 2001-01-02, rainfall = 0.0
ID = 1, date = 2001-01-03, rainfall = 0.0
ID = 1, date = 2001-01-04, rainfall = 0.0
ID = 1, date = 2001-01-05, rainfall = 0.0
ID = 1, date = 2001-01-06, rainfall = 0.0
ID = 1, date = 2001-01-07, rainfall = 0.0

ID = 2, date = 2001-01-01, rainfall = 0.0
ID = 2, date = 2001-01-04, rainfall = 0.0
ID = 2, date = 2001-01-05, rainfall = 0.0
ID = 2, date = 2017-01-01, rainfall = 0.0  # Not within the desired period, so is excluded
ID = 2, date = 2017-01-02, rainfall = 0.0  # Not within the desired period, so is excluded
ID = 2, date = 2017-01-03, rainfall = 0.0  # Not within the desired period, so is excluded
ID = 2, date = 2017-01-04, rainfall = 0.0  # Not within the desired period, so is excluded

如果我搜索2001-01-01和2001-01-07之间数据且所有数据值至少为6的所有位置，则只返回位置1（ID = 1）。不应返回第二个位置（ID = 2），因为它在所需的时间段内没有所需数量的值。

Answer 1

这是我的答案的新版本，因为我误解了“跨越＆＃39;”。我对这个问题的解释仍然是，在计算数值时，我们应该只在给定的时间内完成它

SELECT l.id, l.name 
FROM location l
JOIN location_data d 
ON l.id = d.id 
GROUP BY l.id, l.name
HAVING MIN(d.date) <= '2017-01-01' 
   AND MAX(d.date) >= '2017-12-31'  
   AND (SELECT COUNT(b.id)
        FROM location_data b
        WHERE b.date BETWEEN '2017-01-01' AND '2017-12-31' 
        AND b.id = l.id) >= 2

我将表数据重命名为test db中的location_data，但这可能很明显：）

Answer 2

首先，您应该将日期等值传递给参数，而不是字符串。其次，COUNT()返回一个数字，因此比较应该是数字，而不是字符串。

您应该可以使用group by和having执行所需操作。一种方法是：

SELECT l.ID, l.name 
FROM location l JOIN
     data d
     ON d.id = l.id 
GROUP BY l.id, l.name
HAVING MIN(d.date) <= '$start_date' AND
       MAX(d.date) >= '$end_date' AND
       COUNT(*) >= $min
ORDER BY l.ID;

SQL选择COUNT（*）＆gt;最低人数

2 个答案: