用 UNION ALL 来做减法汇总试试？

SELECT SUM(CNT)
FROM (SELECT COUNT(1) AS CNT
FROM AC02_TEMP AS A
WHERE A.AAB301 IN (SELECT AAB301 FROM AA26 WHERE AAA148 = '130800')
AND A.AAE200 = '41'
AND A.AAC031 = '1'
AND A.AAC030 >= '2018-01-01 00:00:00'

UNION ALL
SELECT -1 * COUNT(1) AS CNT
FROM AC02_TEMP AS A
WHERE A.AAB301 IN (SELECT AAB301 FROM AA26 WHERE AAA148 = '130800')
AND A.AAE200 = '41'
AND A.AAC031 = '1'
AND A.AAC030 >= '2018-01-01 00:00:00'
AND EXISTS (SELECT FROM AC02_TEMP AS B
WHERE B.AAC001 = A.AAC001
AND B.AAC030 < '2018-01-01 00:00:00'));

@sss15 感谢您的回复，首先连接处应该是 ON t0.TID = A.AAC001 ，然后，AAC001 是这张表的联合主键之一，类型为 Int64 ，未关联上时值不是 NULL 而是 0 （这一点我也不知道为什么要这么实现，按理说未关联上不应该是 NULL 吗），然后我执行了 该查询，依然很慢
```
EXPLAIN indexes = 1 SELECT
COUNT( 1 ) AS "新参保人数"
FROM
AC02_TEMP AS A
LEFT JOIN ( SELECT B.AAC001 AS TID FROM AC02_TEMP AS B WHERE B.AAC030 < '2018-01-01 00:00:00' ) AS t0 ON t0.TID = A.AAC001
WHERE
A.AAB301 IN ( SELECT AAB301 FROM AA26 WHERE AAA148 = '130800' )
AND A.AAE200 = '41'
AND A.AAC031 = '1'
AND A.AAC030 >= '2018-01-01 00:00:00'
AND t0.TID = 0;
```

执行计划
```
CreatingSets (Create sets before main query execution)
Expression ((Projection + Before ORDER BY))
Aggregating
Expression (Before GROUP BY)
Filter (WHERE)
Join (JOIN FillRightFirst)
Filter (( + Before JOIN))
ReadFromMergeTree (default.AC02_TEMP)
Indexes:
PrimaryKey
Keys:
AAE200
" Condition: (AAE200 in ['41', '41'])"
Parts: 2/2
Granules: 4821/4821
Expression ((Joined actions + (Rename joined columns + (Projection + Before ORDER BY))))
ReadFromMergeTree (default.AC02_TEMP)
Indexes:
PrimaryKey
Condition: true
Parts: 2/2
Granules: 4821/4821
CreatingSet (Create set for subquery)
Expression ((Projection + Before ORDER BY))
ReadFromMergeTree (default.AA26)
Indexes:
PrimaryKey
Condition: true
Parts: 1/1
Granules: 1/1

```

不懂 clickhouse ，硬要一个 SQL 出来结果的话，这个行不行，按日期分两段去重查 KEY ，再统计，最后去除重复的，这样也没有大数据集匹配

SELECT COUNT(1)
FROM (SELECT AAC001, COUNT(1) AS cnt
FROM (SELECT DISTINCT AAC001
FROM AC02_TEMP AS A
WHERE A.AAB301 IN
(SELECT AAB301 FROM AA26 WHERE AAA148 = '130800')
AND A.AAE200 = '41'
AND A.AAC031 = '1'
AND A.AAC030 >= '2018-01-01 00:00:00'
UNION ALL
SELECT DISTINCT AAC001
FROM AC02_TEMP AS A
WHERE A.AAB301 IN
(SELECT AAB301 FROM AA26 WHERE AAA148 = '130800')
AND A.AAE200 = '41'
AND A.AAC031 = '1'
AND A.AAC030 < '2018-01-01 00:00:00'))
GROUP BY AAC001)
WHERE CNT = 1;

看场景，你是要一次性导出数据还是在业务系统里需要动态查询？

如果是一次性导出数据，直接 dump 整表，然后在 python 里用 set 加载，判断 in

In Oracle we can only put up to 1000 values into an IN clause.
Oracle In 字句最多只能包含 1000 个值，否则影响性能。

解决方案：
1 ，用 subquery 把数据集分开处理；
2 ，创建临时表或者物理表来存储中间数据集；
3 ，用 join 来替代 in

@sunrealzhang 类型为 Int64 ，未关联上时值不是 NULL 而是 0 （这一点我也不知道为什么要这么实现，按理说未关联上不应该是 NULL 吗）

这是 ck 的特性，未关联上的，会根据字段类型返回默认值，int 是 0 ，string 是''，其他的没研究