mysql - 如何分组并计算组的平均值

Question

我有以下简单的风表，由控制风传感器的嵌入式设备填充。

describe wind;

Field        Type      Null   Key    Default    Extra
average      float     YES    NULL
min          float     YES    NULL
max          float     YES    NULL
direction    int(11)   YES    NULL
timestamp    datetime  YES    NULL

方向以度为单位，范围从 0 到 360。

典型条目:

select * from wind order by timestamp desc limit 5;
average    min    max    direction    timestamp
7.34167    5.9    10.2    307        2018-04-26 10:24:07
6.655      4.2    8.9     301        2018-04-26 10:23:03
8.74667    7.2    10.8    307        2018-04-26 10:21:59
9.4925     7.4    10.7    295        2018-04-26 10:20:55
11.7175    8.6    14.2    306        2018-04-26 10:19:51

我想计算沿 N、NNE 等 16 个风轴的平均风分布。

基于此group by range in mysql我能够进行分组。 (请注意，我不能使用“下限”解决方案，因为北区间范围为 348.25->360 AND 0->11.25)

我想我已经差不多了，但是我找不到正确的语法来求平均值(实际上是按风向分组的AVG(平均值))。

这是我当前的代码:

select w.dir as `wind direction`, count(*) as `occurence`
from (
    select CASE
    when direction BETWEEN 11.25 and 33.75 then 'NNE'
    when direction BETWEEN 33.75 and 56.25 then 'NE'
    when direction BETWEEN 56.25 and 78.75 then 'ENE'
    when direction BETWEEN 78.75 and 101.25 then 'E'
    when direction BETWEEN 101.25 and 123.75 then 'ESE'
    when direction BETWEEN 123.75 and 146.25 then 'SE'
    when direction BETWEEN 146.25 and 168.75 then 'SSE'
    when direction BETWEEN 168.75 and 191.25 then 'S'
    when direction BETWEEN 191.25 and 213.75 then 'SSW'
    when direction BETWEEN 213.75 and 236.25 then 'SW'
    when direction BETWEEN 236.25 and 258.75 then 'WSW'
    when direction BETWEEN 258.75 and 281.25 then 'W'
    when direction BETWEEN 281.25 and 303.75 then 'WNW'
    when direction BETWEEN 303.75 and 325.75 then 'NW'
    when direction BETWEEN 325.75 and 348.25 then 'NNW'
    else 'N'
     end as `dir`
    from wind) w
    group by w.dir

我尝试添加

 AVG(w.average) as `mean`

在第一行，但我收到错误#1054。我不确定如何包含平均列以便能够计算其平均值。

<小时/>

奖励:我还有一个次要问题:分组是按字母顺序完成的，如下所示:

wind direction    occurence
E                 31
ENE               58
ESE               66
N                 212
NE                128
NNE               62
NNW               326
NW                449
S                 108
SE                133
SSE               192
SSW               355
SW                47
W                 173
WNW               333
WSW               22

我想保持我的案例陈述的顺序(理想情况下我希望 North N 案例成为第一个)。

order by w.dir

做同样的事情，按字母顺序排序。如何保持案件秩序？

非常感谢

Answer 1

我认为这个查询应该给你你想要的结果(没有足够的样本数据来确定)。基本上你只需要在子查询中选择平均值和方向，然后你就可以获得你想要的平均值，并且你可以按原始方向数排序(我们添加 11.75 并使其模 360，以便方向从 0 开始) N，NNW 可达 359)。

select w.dir as `wind direction`, round(avg(average),4) as mean, count(*) as `occurence`
from (
    select CASE
    when direction BETWEEN 11.25 and 33.75 then 'NNE'
    when direction BETWEEN 33.75 and 56.25 then 'NE'
    when direction BETWEEN 56.25 and 78.75 then 'ENE'
    when direction BETWEEN 78.75 and 101.25 then 'E'
    when direction BETWEEN 101.25 and 123.75 then 'ESE'
    when direction BETWEEN 123.75 and 146.25 then 'SE'
    when direction BETWEEN 146.25 and 168.75 then 'SSE'
    when direction BETWEEN 168.75 and 191.25 then 'S'
    when direction BETWEEN 191.25 and 213.75 then 'SSW'
    when direction BETWEEN 213.75 and 236.25 then 'SW'
    when direction BETWEEN 236.25 and 258.75 then 'WSW'
    when direction BETWEEN 258.75 and 281.25 then 'W'
    when direction BETWEEN 281.25 and 303.75 then 'WNW'
    when direction BETWEEN 303.75 and 325.75 then 'NW'
    when direction BETWEEN 325.75 and 348.25 then 'NNW'
    else 'N'
     end as `dir`,
    average,
    direction
    from wind) w
group by w.dir
order by floor(w.direction + 11.75) mod 360