Question

我正在使用以下查询：

SELECT p.bp_pod, COUNT(j.max) as 'COUNT',
SUM (CASE WHEN j.max <= 120 AND j.min <= 80 THEN 1 ELSE 0 END),
SUM (CASE WHEN  j.max > 120 AND j.max < 140 OR j.min > 80 AND j.min < 90 THEN 1 ELSE 0 END),
SUM (CASE WHEN  j.max >= 140 OR j.min >= 90 THEN 1 ELSE 0 END)
FROM (SELECT 0 as bp_pod UNION ALL
SELECT 1 as bp_pod UNION ALL
SELECT 2 as bp_pod )
p LEFT JOIN
tb_journal j
ON p.bp_pod = j.bp_pod
GROUP BY p.bp_pod LIMIT 0, 90;

但是此查询根据bp_pod对记录进行分组，这些记录只能包含三个值：0, 1, 2。

以下是我正在使用的table

CREATE TABLE "table" ( `bp_pod` NUMERIC, `max` NUMERIC, `min` NUMERIC )

以下是我正在使用的一些值：

INSERT INTO table VALUES('21', '134', '78');
INSERT INTO table VALUES('13', '126', '88');
INSERT INTO table VALUES('6', '97', '54');

现在我更改了应用程序并且它可以有多个值，因此我需要使用范围实现GROUP BY，例如：

bp_pod > 5 AND bp_pod < 12
bp_pod >= 12 AND bp_pod < 21
bp_pod >= 21 AND bp_pod < 24

但在索引pod 0,1,2。

中包含此类范围

预期结果如下：

pod    COUNT   range_1 range_2 range_3
0       0       1        0       0
1       0       0        1       0
2       0       0        0       1

有任何线索吗？谢谢。

Answer 1

由于left join，您需要小心。我认为这样做你想要的：

SELECT p.pod,
       COUNT(j.bp_max) as 'COUNT',
       SUM(CASE WHEN j.max <= 120 AND j.min <= 80 THEN 1 ELSE 0 END),
       SUM(CASE WHEN j.max > 120 AND j.max < 140 OR j.min > 80 AND j.min < 90 THEN 1 ELSE 0 END),
       SUM(CASE WHEN j.max >= 140 OR j.min >= 90 THEN 1 ELSE 0 END)
FROM (SELECT 0 as bp_pod, 5 as minb, 12 as maxb UNION ALL
      SELECT 1 as bp_pod, 12, 21 UNION ALL
      SELECT 2 as bp_pod, 21, 24
     ) p LEFT JOIN
     tb_journal j
     ON j.bp_pod >= p.minb AND j.bp_pod < p.maxb
GROUP BY p.bp_pod;

SQLite，GROUP BY和RANGE

1 个答案: