Question

我的数据格式如下：

date（mm / dd / yyyy），desc（详情desc），category（a，b，c），tran_type（借记卡，赠送）， amount spent（金额）。

我想以下列格式获取数据：

category  ||  tran_type   ||   Jan_total  ||  feb_total  ||  mar_total  
A         ||  debit       ||   $101       ||  $201       ||  $302  
A         ||  credit      ||   $500       ||  $600       ||  $200

Answer 1

此查询应该为您提供所需的结果。

SELECT category
  ,tran_type
  ,SUM(IF(month(date) = 1,ABS(amount),0)) as jan_total
  ,SUM(IF(month(date) = 2,ABS(amount),0)) as feb_total
  ,SUM(IF(month(date) = 3,ABS(amount),0)) as mar_total
  ,SUM(IF(month(date) = 4,ABS(amount),0)) as apr_total
  ,SUM(IF(month(date) = 5,ABS(amount),0)) as may_total
  ,SUM(IF(month(date) = 6,ABS(amount),0)) as jun_total
  ,SUM(IF(month(date) = 7,ABS(amount),0)) as jul_total
  ,SUM(IF(month(date) = 8,ABS(amount),0)) as aug_total
  ,SUM(IF(month(date) = 9,ABS(amount),0)) as sep_total
  ,SUM(IF(month(date) = 10,ABS(amount),0)) as okt_total
  ,SUM(IF(month(date) = 11,ABS(amount),0)) as nov_total
  ,SUM(IF(month(date) = 12,ABS(amount),0)) as dec_total
 FROM transactions
 WHERE YEAR(date) = '2011'
 GROUP BY category, tran_type

如果您不想遇到麻烦，请不要忘记过滤年份。

Answer 2

那几年呢？您是否希望Janury专栏在2011年1月添加2011年1月的交易？

我假设您需要单独添加Year列或为其使用WHERE子句，因此包含该示例的两个选项：

基本上你需要生成12个子查询才能实现这一点，而SQL看起来会很混乱，并不是实现这个目标的常用方法：

SELECT category, YEAR(date), tran_type,

    (SELECT SUM(amounts) FROM TableName s1 WHERE YEAR(t.date)=YEAR(s1.date) AND MONTH(s1.date)=1 AND t.category=s1.category AND t.tran_type=s1.tran_type) AS 'Jan_Total',

    (SELECT SUM(amounts) FROM TableName s2 WHERE YEAR(t.date)=YEAR(s2.date) AND MONTH(s2.date)=2 AND t.category=s2.category AND t.tran_type=s2.tran_type) AS 'Feb_Total',

     ....REPEAT ABOVE 2 LINES FOR EACH MONTH

FROM TableName t
WHERE t.date>'2011-01-01'
GROUP BY t.category, YEAR(t.date), t.tran_type;

如上所述，上述内容并不完全优雅，更好的解决方案是使用以下SQL并将表示层中的数据格式化为用户：

SELECT category, YEAR(date), MONTH(date), tran_type, SUM(amounts)
FROM TableName
GROUP BY TableName;

希望有所帮助。

Answer 3

单个月（2011年1月）您可以使用：

SELECT category, tran_type, SUM(amount_spent) AS total FROM myTable WHERE date>="2011-01-01" AND date<"2011-02-01" GROUP BY category, tran_type;

假设您的索引至少为date，那么这应该相当快。更好的索引是(date, category, tran_type)，因为这也有助于分组。

如果您希望以您描述的格式组合多个月，您可以将这些查询组合为子查询，例如：

SELECT jan.category, jan.tran_type, jan.total, feb.total FROM
    (SELECT category, tran_type, SUM(amount_spent) AS total FROM myTable WHERE date>="2011-01-01" AND date<"2011-02-01" GROUP BY category, tran_type) AS jan,
    (SELECT category, tran_type, SUM(amount_spent) AS total FROM myTable WHERE date>="2011-02-01" AND date<"2011-03-01" GROUP BY category, tran_type) AS feb
 WHERE jan.category = feb.category
 AND jan.tran_type = feb.tran_type;

虽然我怀疑获取您想要的数据的最简单方法（虽然不是非常有效且不是您所描述的格式），但是：

SELECT category, tran_type, MONTH(date) AS theMonth, SUM(amount_spent) AS total FROM myTable WHERE date>="2011-01-01" AND date<"2012-01-01" GROUP BY category, tran_type, theMonth;

MONTH()函数的使用远非最佳，但如果您不运行通常可能的查询。

mySQL查询列中的总和（月）

3 个答案: