Question

我有一个包含帐户和订阅金额的表格，或多或少有以下查询：

SELECT FORMAT_DATE(\"%Y-%m\", DATE(receivedOn)) AS Date,
       SUM(amount)
FROM `mytablehere`
GROUP BY Date
ORDER BY Date ASC

输出很好但是有些月份没有任何订阅，所以金额为空，并且最终结果中没有返回月份。由于我直接在Google表格中导入结果以从中构建图表，因此没有整整12个月的年份令人讨厌。

有没有办法生成＆＃34;虚拟＆＃34;错过月份的月份？回答一个类似的问题建议创建另一个表满月并联合两个表...我正在寻找更简单的东西...

Answer 1

有没有办法生成＆＃34;虚拟＆＃34;错过月份的月份？

下面的示例适用于BigQuery Standard SQL

#standardSQL
WITH `mytablehere` AS (
  SELECT '2018-01-01' AS receivedOn, 1 amount UNION ALL
  SELECT '2018-01-02' AS receivedOn, 2 amount UNION ALL
  SELECT '2018-03-04' AS receivedOn, 4 amount UNION ALL
  SELECT '2018-05-10' AS receivedOn, 3 amount UNION ALL
  SELECT '2018-05-11', 5
), months AS (
  SELECT month FROM (
    SELECT 
      MIN(PARSE_DATE('%Y-%m-%d', receivedOn)) AS min_month, 
      MAX(PARSE_DATE('%Y-%m-%d', receivedOn)) AS max_month 
    FROM `mytablehere`
  ), UNNEST(GENERATE_DATE_ARRAY(min_month, max_month, INTERVAL 1 MONTH)) month
)
SELECT 
  FORMAT_DATE('%Y-%m', month) AS DATE,
  SUM(amount) amount
FROM months
LEFT JOIN `mytablehere`
ON FORMAT_DATE('%Y-%m', PARSE_DATE('%Y-%m-%d', receivedOn)) = FORMAT_DATE('%Y-%m', month)
GROUP BY DATE
ORDER BY DATE ASC

结果为

Row DATE    amount   
1   2018-01 3    
2   2018-02 null     
3   2018-03 4    
4   2018-04 null     
5   2018-05 8

按月分组＆＃34;虚拟＆＃34;一个月得到一整年

1 个答案: