用于评估某事物“开启”的总持续时间的 MySQL 查询

发布于 2024-08-20 18:40:07 字数 560 浏览 4 评论 0原文

我在表中有一系列带时间戳的开/关数据,代表开/关状态,或者状态“开始”的点,

00:00:00    0
04:00:00    1
08:00:00    0
09:00:00    1
15:00:00    0
20:00:00    1
23:59:59    0

我需要计算(例如)24 小时内开状态的总持续时间。

在此简化示例中,总持续时间 = 1 为 (04:00:00->08:00:00、09:00:00->15:00:00、20:00:00->23:59:59 即 13:59:59 大约 14h

我无法确定这是否可以单独在 SQL 中完成,或者我使用的底层框架 (django) 是否需要根据返回的数据来完成此操作。如果可能的话,我显然更愿意让数据库承担繁重的工作,因为我们可能还需要在单独的统计数据包中使用 SQL。

我不清楚我是否可以对选择中的上一个或下一个元素进行操作,我是一个自信的 SQL 用户,但不知道从哪里开始这个或通用方法,有什么想法吗?

我真的很喜欢在一个查询中做到这一点,或者用其他一些聪明的方法来计算我所缺少的!

I have a series of timestamped on/off data in a table, representing on/off states, or the point at which a state "starts"

00:00:00    0
04:00:00    1
08:00:00    0
09:00:00    1
15:00:00    0
20:00:00    1
23:59:59    0

I need to calculate the total duration of (say) the ON state over a 24h period.

In this simplified example total duration = 1 is
(04:00:00->08:00:00, 09:00:00->15:00:00, 20:00:00->23:59:59
i.e. 13:59:59 approx 14h

I can't determine whether this can be done in SQL alone, or whether the underlying framework i am using (django) would need to do this based on returned data. I would obviously prefer to have the database do the heavy lifting if possible, because we may need to use the SQL in our separate stats package as well.

It's not clear to me if I can do operations on (say) previous or next element in the select, I am a confident SQL user but can't see where to start for this or the generalised approach, any ideas?

I'd really like this in a single query, or some other clever way of calculating this I am missing!

如果你对这篇内容有疑问,欢迎到本站社区发帖提问 参与讨论,获取更多帮助,或者扫码二维码加入 Web 技术交流群。

扫码二维码加入Web技术交流群

发布评论

需要 登录 才能够评论, 你可以免费 注册 一个本站的账号。

评论(1

黎夕旧梦 2024-08-27 18:40:08

MySQL 中没有 row_number(),但您可以执行双连接来搜索上一行:

select 
    sum(case when cur.state = 0 then 0
        else subtime(cur.timeCol, prev.timeCol)
        end) as TotalOnTime
from YourTable cur
join YourTable prev
    on prev.timeCol < cur.timeCol
left join YourTable inbetween
    on prev.timeCol < inbetween.timeCol
    and inbetween.timeCol < cur.timeCol
where inbetween.timeCol is null;

在 MySQL 中,您还可以使用变量,在这种情况下可能更有效:

set @total := '00:00:00';
set @lasttime := '00:00:00';

select 
    @total := addtime(@total, case 
        when state = 0 then 0
        when @lasttime is null then 0
        else subtime(timeCol, @lasttime)
        end)
,   @lasttime := timeCol
from YourTable
order by timeCol;

select 'Result = ', @total;

代码为创建并填充测试表:

DROP TABLE IF EXISTS YourTable;
CREATE TABLE YourTable (
   timeCol time,
   state bit
) ENGINE=MyISAM DEFAULT CHARSET=latin1;

insert into YourTable values ('00:00:00',    0);
insert into YourTable values ('04:00:00',    1);
insert into YourTable values ('08:00:00',    0);
insert into YourTable values ('09:00:00',    1);
insert into YourTable values ('15:00:00',    0);
insert into YourTable values ('20:00:00',    1);
insert into YourTable values ('23:59:59',    0);

There's no row_number() in MySQL, but you can do a double join to search for the previous row:

select 
    sum(case when cur.state = 0 then 0
        else subtime(cur.timeCol, prev.timeCol)
        end) as TotalOnTime
from YourTable cur
join YourTable prev
    on prev.timeCol < cur.timeCol
left join YourTable inbetween
    on prev.timeCol < inbetween.timeCol
    and inbetween.timeCol < cur.timeCol
where inbetween.timeCol is null;

In MySQL, you can also use a variable, which in this case is probably more efficient:

set @total := '00:00:00';
set @lasttime := '00:00:00';

select 
    @total := addtime(@total, case 
        when state = 0 then 0
        when @lasttime is null then 0
        else subtime(timeCol, @lasttime)
        end)
,   @lasttime := timeCol
from YourTable
order by timeCol;

select 'Result = ', @total;

Code to create and populate test table:

DROP TABLE IF EXISTS YourTable;
CREATE TABLE YourTable (
   timeCol time,
   state bit
) ENGINE=MyISAM DEFAULT CHARSET=latin1;

insert into YourTable values ('00:00:00',    0);
insert into YourTable values ('04:00:00',    1);
insert into YourTable values ('08:00:00',    0);
insert into YourTable values ('09:00:00',    1);
insert into YourTable values ('15:00:00',    0);
insert into YourTable values ('20:00:00',    1);
insert into YourTable values ('23:59:59',    0);
~没有更多了~
我们使用 Cookies 和其他技术来定制您的体验包括您的登录状态等。通过阅读我们的 隐私政策 了解更多相关信息。 单击 接受 或继续使用网站,即表示您同意使用 Cookies 和您的相关数据。
原文