两个SELECT语句作为两列

时间:2016-01-18 15:21:44

标签: sql postgresql

这似乎是一个简单的查询,但我正在努力解决它。

以下是我的数据样本。

user_id  dated
463      2016-01-01
463      2016-01-02
1456     2016-01-01
1456     2016-01-02
1398     2015-12-01
1398     2015-12-02

我希望在两个不同的时间段内获得唯一身份用户的数量。以下是我想从单行和两列中获取组合输出的查询。

-- 60
SELECT COUNT(DISTINCT(tld.user_id)) count_active_users_60
FROM table tld
WHERE tld.dated BETWEEN (NOW() - INTERVAL '60 days') AND (NOW() - INTERVAL '30 days')

-- 30
SELECT COUNT(DISTINCT(tld.user_id)) count_active_users_30
FROM table tld
WHERE tld.dated >= NOW() - INTERVAL '30 days'

我想要一个看起来像这样的输出:

count_active_users_60    count_active_users_30
1                        2

我一直在搞乱各种CASE语句和子选择,但是不同的条款让我失望。

SELECT COUNT(DISTINCT(rar.user_id))
FROM 
(
SELECT user_id,
COUNT(CASE WHEN tld.dated BETWEEN (NOW() - INTERVAL '60 days') AND (NOW() - INTERVAL '30 days') THEN 1 ELSE NULL END) AS count_active_users_60,
COUNT(CASE WHEN tld.dated >= NOW() - INTERVAL '30 days' THEN 1 ELSE NULL END) AS count_active_users_30
FROM testing_login_duration tld
GROUP BY user_id
) rar;

1 个答案:

答案 0 :(得分:3)

使用条件聚合:

SELECT COUNT(DISTINCT CASE WHEN tld.dated BETWEEN (NOW() - INTERVAL '60 days') AND (NOW() - INTERVAL '30 days')
                           THEN tld.user_id
                       END) count_active_users_60,
       COUNT(DISTINCT CASE WHEN tld.dated >= NOW() - INTERVAL '30 days'
                           THEN tld.user_id
                       END) count_active_users_30
FROM table tld
WHERE tld.dated >= NOW() - INTERVAL '60 days';