Question

下面我有示例数据，我想将所有 biggest_win 和乘数合并到玩家1083987121的一列中！

select * from player;

player_key, country, biggest_win, multiplier, date
1083987121, USD, 490.2, 163.4, 2020-03-17
1083987121, USD, 256.0, 128.0, 2020-03-17
1083987121, USD, 210.0, 70.0, 2020-03-17

合并后的期望

1083987121, USD, 490.2#256.0#256.0 163.4#128.0#70.0, 2020-03-17
1083987121, USD, 490.2#256.0#256.0 163.4#128.0#70.0, 2020-03-17
1083987121, USD, 490.2#256.0#256.0 163.4#128.0#70.0, 2020-03-17

有人对此有好的解决方案吗？我已经尝试过XML函数，在BigQuery上似乎不起作用

Answer 1

这回答了问题的原始版本。

select player_key, max(country) as country,
       array_agg(biggest_win order by biggest_win desc),
       array_agg(multiplier order by biggest_win desc),
       max(date) as date
from t
group by player_key;

注意：这将返回数组而不是字符串中的值-数组更易于使用。但是，您也可以使用string_agg()：

select player_key, max(country) as country,
       string_agg(biggest_win, '#' order by biggest_win desc),
       string_agg(multiplier, '#' order by biggest_win desc),
       max(date) as date
from t
group by player_key, country;

Answer 2

以下是BigQuery标准SQL

#standardSQL
SELECT player_key, country, 
  STRING_AGG(CAST(biggest_win AS STRING), '#') biggest_win, 
  STRING_AGG(CAST(multiplier AS STRING), '#') multiplier, 
  date 
FROM `project.dataset.player`
GROUP BY player_key, country, date

有输出

Row player_key  country biggest_win     multiplier      date     
1   1083987121  USD     490.2#256#210   163.4#128#70    2020-03-17

注意：根据您的I'm expecting after merge示例-我假设您希望合并后的结果-玩家，国家/地区和日期

我想知道是否要打印所有行

如果您仍要输出所有行（在其中看不到太多的值，但请按您的要求），请在下面使用

#standardSQL
SELECT player_key, country, 
  STRING_AGG(CAST(biggest_win AS STRING), '#') OVER(PARTITION BY player_key, country, date) biggest_win, 
  STRING_AGG(CAST(multiplier AS STRING), '#') OVER(PARTITION BY player_key, country, date) multiplier, 
  date 
FROM `project.dataset.player`

在这种情况下，相应的输出是

Row player_key  country biggest_win     multiplier      date     
1   1083987121  USD     490.2#256#210   163.4#128#70    2020-03-17   
2   1083987121  USD     490.2#256#210   163.4#128#70    2020-03-17   
3   1083987121  USD     490.2#256#210   163.4#128#70    2020-03-17

要将这两列合并为一列-在下面使用

#standardSQL
SELECT player_key, country, 
  CONCAT(
    STRING_AGG(CAST(biggest_win AS STRING), '#') OVER(PARTITION BY player_key, country, DATE), ' ', 
    STRING_AGG(CAST(multiplier AS STRING), '#') OVER(PARTITION BY player_key, country, DATE) 
  ) combined_column, 
  DATE 
FROM `project.dataset.player`

有输出

Row player_key  country combined_column             date     
1   1083987121  USD     490.2#256#210 163.4#128#70  2020-03-17   
2   1083987121  USD     490.2#256#210 163.4#128#70  2020-03-17   
3   1083987121  USD     490.2#256#210 163.4#128#70  2020-03-17

如何在BigQuery的一列/行中合并多行值

2 个答案: