Question

CREATE TABLE CDT (
    UID INTEGER NOT NULL
    , COUNTRY varchar(100) NOT NULL
    , MSDN varchar(100) NOT NULL
    , MSSI varchar(100) NOT NULL
    , CALL_DURATION INTEGER NOT NULL
    , SDATA INTEGER NOT NULL
    , COST INTEGER NOT NULL
    , DATE_KEY INTEGER NOT NULL
);

查询以获取数据：

SELECT
    UID
    , COUNTRY
    , MSDN
    , MSSI
    , CALL_DURATION
    , SDATA
    , COST
    , c.DATE_KEY
    , d.DATE_VAL
FROM CDT as C
INNER JOIN DATE_DIMENSION as d on c.DATE_KEY = d.DATE_KEY;

OP：我需要以下计数：每_DAY_COUNT | WEEK_WIE_COUNT | MONTHLY_COUNT | Year_COUNT个

Answer 1

Sholk-下次将您的数据显示为文本而不是图片，以便我们将其复制粘贴到示例中。这样，我花了15分钟才能将示例数据汇总在一起...

你是这个意思吗？

-我设法简化了很多，想到了...

-- your input data
WITH cdt(uid,country,msdn,mssi,call_dur,sdata,cost,date_key,date_val) AS (
            SELECT 50,'IND',9861388009,9761588009,466,477,48900,3290,DATE '2019-01-04'
  UNION ALL SELECT 51,'IND',9861388009,9761588009,466,477,48900,3290,DATE '2019-01-04'
  UNION ALL SELECT 52,'IND',9861388009,9761588009,466,477,48900,3290,DATE '2019-01-04'
  UNION ALL SELECT 53,'IND',9861388009,9761588009,466,477,48900,3290,DATE '2019-01-04'
  UNION ALL SELECT 45,'IND',9861388009,9761588009,466,477,48900,3291,DATE '2019-01-05'
  UNION ALL SELECT 46,'IND',9861388009,9761588009,466,477,48900,3291,DATE '2019-01-05'
  UNION ALL SELECT 47,'IND',9861388009,9761588009,466,477,48900,3291,DATE '2019-01-05'
  UNION ALL SELECT 48,'IND',9861388009,9761588009,466,477,48900,3291,DATE '2019-01-05'
  UNION ALL SELECT 41,'IND',9861388009,9761588009,466,477,48900,3292,DATE '2019-01-06'
  UNION ALL SELECT 42,'IND',9861388009,9761588009,466,477,48900,3292,DATE '2019-01-06'
  UNION ALL SELECT 43,'IND',9861388009,9761588009,466,477,48900,3292,DATE '2019-01-06'
  UNION ALL SELECT 44,'IND',9861388009,9761588009,466,477,48900,3292,DATE '2019-01-06'
  UNION ALL SELECT 37,'IND',9861388009,9761588009,466,477,48900,3293,DATE '2019-01-07'
  UNION ALL SELECT 38,'IND',9861388009,9761588009,466,477,48900,3293,DATE '2019-01-07'
  UNION ALL SELECT 39,'IND',9861388009,9761588009,466,477,48900,3293,DATE '2019-01-07'
  UNION ALL SELECT 40,'IND',9861388009,9761588009,466,477,48900,3293,DATE '2019-01-07'
  UNION ALL SELECT 33,'IND',9861388009,9761588009,466,477,48900,3294,DATE '2019-01-08'
  UNION ALL SELECT 34,'IND',9861388009,9761588009,466,477,48900,3294,DATE '2019-01-08'
  UNION ALL SELECT 35,'IND',9861388009,9761588009,466,477,48900,3294,DATE '2019-01-08'
  UNION ALL SELECT 36,'IND',9861388009,9761588009,466,477,48900,3294,DATE '2019-01-08'
  UNION ALL SELECT 29,'IND',9861388009,9761588009,466,477,48900,3295,DATE '2019-01-09'
)
-- real query starts here ...
SELECT
  date_val
, COUNT(*) OVER(PARTITION BY DAY  (date_val)) AS per_day
, COUNT(*) OVER(PARTITION BY WEEK (date_val)) AS per_week
, COUNT(*) OVER(PARTITION BY MONTH(date_val)) AS per_month
, COUNT(*) OVER(PARTITION BY YEAR (date_val)) AS per_year
FROM cdt
;

   date_val  | per_day | per_week | per_month | per_year 
 ------------+---------+----------+-----------+----------
  2019-01-05 |       4 |        8 |        21 |       21
  2019-01-05 |       4 |        8 |        21 |       21
  2019-01-05 |       4 |        8 |        21 |       21
  2019-01-05 |       4 |        8 |        21 |       21
  2019-01-04 |       4 |        8 |        21 |       21
  2019-01-04 |       4 |        8 |        21 |       21
  2019-01-04 |       4 |        8 |        21 |       21
  2019-01-04 |       4 |        8 |        21 |       21
  2019-01-09 |       1 |       13 |        21 |       21
  2019-01-08 |       4 |       13 |        21 |       21
  2019-01-08 |       4 |       13 |        21 |       21
  2019-01-08 |       4 |       13 |        21 |       21
  2019-01-08 |       4 |       13 |        21 |       21
  2019-01-07 |       4 |       13 |        21 |       21
  2019-01-07 |       4 |       13 |        21 |       21
  2019-01-07 |       4 |       13 |        21 |       21
  2019-01-07 |       4 |       13 |        21 |       21
  2019-01-06 |       4 |       13 |        21 |       21
  2019-01-06 |       4 |       13 |        21 |       21
  2019-01-06 |       4 |       13 |        21 |       21
  2019-01-06 |       4 |       13 |        21 |       21
 (21 rows)

Vertica SQL for Data以列，日，弱，月和年为列排列

1 个答案: