从日期维度 table 到层次结构

Question

这是我的 table:

idDate  timeformat  timeformatdate idYear YearName idSemester semestername idquarter quartername idmonth month idweek week idday day  

20160101    2016-01-01  01-Jan-16   2016    2016    1   S1  1   Q1  201601  Jan         53  W53         1   Friday   , 1                                      
20160102    2016-01-02  02-Jan-16   2016    2016    1   S1  1   Q1  201601  Jan         53  W53         2   Saturday , 2                                      
20160103    2016-01-03  03-Jan-16   2016    2016    1   S1  1   Q1  201601  Jan         53  W53         3   Sunday   , 3                                      
20160104    2016-01-04  04-Jan-16   2016    2016    1   S1  1   Q1  201601  Jan         1   W1          4   Monday   , 4                                      
20160105    2016-01-05  05-Jan-16   2016    2016    1   S1  1   Q1  201601  Jan         1   W1          5   Tuesday  , 5                                      
20160106    2016-01-06  06-Jan-16   2016    2016    1   S1  1   Q1  201601  Jan         1   W1          6   Wednesday, 6                                      
20160107    2016-01-07  07-Jan-16   2016    2016    1   S1  1   Q1  201601  Jan         1   W1          7   Thursday , 7                                      
20160108    2016-01-08  08-Jan-16   2016    2016    1   S1  1   Q1  201601  Jan         1   W1          8   Friday   , 8                                      
20160109    2016-01-09  09-Jan-16   2016    2016    1   S1  1   Q1  201601  Jan         1   W1          9   Saturday , 9                                      
20160110    2016-01-10  10-Jan-16   2016    2016    1   S1  1   Q1  201601  Jan         1   W1          10  Sunday   , 10                                     
20160111    2016-01-11  11-Jan-16   2016    2016    1   S1  1   Q1  201601  Jan         2   W2          11  Monday   , 11                                     
20160112    2016-01-12  12-Jan-16   2016    2016    1   S1  1   Q1  201601  Jan         2   W2          12  Tuesday  , 12                                     
20160113    2016-01-13  13-Jan-16   2016    2016    1   S1  1   Q1  201601  Jan         2   W2          13  Wednesday, 13                                     
20160114    2016-01-14  14-Jan-16   2016    2016    1   S1  1   Q1  201601  Jan         2   W2          14  Thursday , 14                                     
20160115    2016-01-15  15-Jan-16   2016    2016    1   S1  1   Q1  201601  Jan         2   W2          15  Friday   , 15                                     
20160116    2016-01-16  16-Jan-16   2016    2016    1   S1  1   Q1  201601  Jan         2   W2          16  Saturday , 16                                     
20160117    2016-01-17  17-Jan-16   2016    2016    1   S1  1   Q1  201601  Jan         2   W2          17  Sunday   , 17                                     
20160118    2016-01-18  18-Jan-16   2016    2016    1   S1  1   Q1  201601  Jan         3   W3          18  Monday   , 18                                     
20160119    2016-01-19  19-Jan-16   2016    2016    1   S1  1   Q1  201601  Jan         3   W3          19  Tuesday  , 19                                     
20160120    2016-01-20  20-Jan-16   2016    2016    1   S1  1   Q1  201601  Jan         3   W3          20  Wednesday, 20                                     
20160121    2016-01-21  21-Jan-16   2016    2016    1   S1  1   Q1  201601  Jan         3   W3          21  Thursday , 21                                     
20160122    2016-01-22  22-Jan-16   2016    2016    1   S1  1   Q1  201601  Jan         3   W3          22  Friday   , 22                                     
20160123    2016-01-23  23-Jan-16   2016    2016    1   S1  1   Q1  201601  Jan         3   W3          23  Saturday , 23                                     
20160124    2016-01-24  24-Jan-16   2016    2016    1   S1  1   Q1  201601  Jan         3   W3          24  Sunday   , 24                                     
20160125    2016-01-25  25-Jan-16   2016    2016    1   S1  1   Q1  201601  Jan         4   W4          25  Monday   , 25                                     
20160126    2016-01-26  26-Jan-16   2016    2016    1   S1  1   Q1  201601  Jan         4   W4          26  Tuesday  , 26                                     
20160127    2016-01-27  27-Jan-16   2016    2016    1   S1  1   Q1  201601  Jan         4   W4          27  Wednesday, 27                                     
20160128    2016-01-28  28-Jan-16   2016    2016    1   S1  1   Q1  201601  Jan         4   W4          28  Thursday , 28                                     
20160129    2016-01-29  29-Jan-16   2016    2016    1   S1  1   Q1  201601  Jan         4   W4          29  Friday   , 29                                     
20160130    2016-01-30  30-Jan-16   2016    2016    1   S1  1   Q1  201601  Jan         4   W4          30  Saturday , 30                                     
20160131    2016-01-31  31-Jan-16   2016    2016    1   S1  1   Q1  201601  Jan         4   W4          31  Sunday   , 31

我正在尝试通过 SQL 语句创建 time_hierarchy。我想要实现的是例如：

Time
[2016] as Year
[Q1, 2016] as Quarter
[Jan, 2016] as Month
[Jan, 1, 2016] as Day
Etc...

我真的不知道如何实现这个，或者如果我的 table 不支持这种层次结构。

你能帮帮我吗？

谢谢

Answer 1

也许是这样的？

SELECT
  CASE WHEN quarterName IS NULL THEN 'Year'
       WHEN month       IS NULL THEN 'Quarter'
       WHEN idDay       IS NULL THEN 'Month'
                                ELSE 'Day'
  END
    AS aggregateLevel,
  CASE WHEN quarterName IS NULL THEN YearName
       WHEN month       IS NULL THEN QuarterName || ', ' || YearName
       WHEN idDay       IS NULL THEN Month || ', ' || YearName
                                ELSE Month || ', ' || CAST(idDay AS VARCHAR(2)) || YearName
  END
    AS periodTitle,
  MIN(idDate)   AS periodFirstDate,
  MAX(idDate)   AS periodFinalDate
FROM
  yourTable
GROUP BY
  GROUPING SETS (
    (YearName),
    (YearName, quarterName),
    (YearName, quarterName, month),
    (YearName, quarterName, month, idDay)
  )
ORDER BY
  MIN(idDate),
  MAX(idDate) DESC

从日期维度 table 到层次结构

From date dimension table to hierarchy

sql

data-warehouse

hierarchy

postgresql-10