sql GROUP BY，GROUPING SETS，ROLLUP，CUBE，GROUPING_ID
Posted
tags:
篇首语：本文由小常识网(cha138.com)小编为大家整理，主要介绍了sql GROUP BY，GROUPING SETS，ROLLUP，CUBE，GROUPING_ID相关的知识，希望对你有一定的参考价值。
/*
Grouping allows you take rows and combine them based on shared values. There are different types of grouping:
GROUP BY - collapses rows that have a shared value into a single row - additional columns that have multiple rows that are collapsed into one must use an aggregate value. For example, if you GROUP BY CustomerID and that CustomerID originally had two orders, for the OrderPrice column you have to use some aggregate value such as SUM, COUNT, MAX, or AVG.
GROUPING SETS - an extention of GROUP BY that allows multiple grouping, good for things like generating grand total. When a NULL comes up in one of the cells, that NULL could be there because it's grouping multiple values (though it might all just be a NULL that's present in your data). GROUPING SETS is how you can identify how you want to group everything, where there are some shortcut groupings that do common grouping tasks like ROLLUP (shows grouping for a heirarchy) and CUBE (which groups by all possible grouping combinations).
ROLLUP - provides a shortcut for defining groups sets with combinations that assume input columns from a hierarchy. Will give grand totals and subtotals for combinations.
CUBE - like ROLLUP, but shows every possible combination. Provides a shortcut for defining goruping sets in which all possible combinations of grouping sets created. This can be good if you're not sure which grouping you want to use, since it'll show you everything (scortched earth grouping).
Finally, in summary rows, GROUPING SETS, ROLLUP, and CUBE will return a NULL value in some columns to show that this is a summary row (for example, if it's showing totals for the Category column, the Category will be listed in that row but all other columns will show NULL, showing you that this is the summary for that Category). This NULL can be indistinguishable from a NULL in the data set. Therefore, the GROUPING_ID function tells you if the NULL is due to a summary row or if there's actually a NULL in the data.
*/
--------------
-- GROUP BY --
--------------
-- GROUP BY - collapses rows that have a shared value into a single row - additional columns that have multiple rows that are collapsed into one must use an aggregate value. For example, if you GROUP BY CustomerID and that CustomerID originally had two orders, for the OrderPrice column you have to use some aggregate value such as SUM, COUNT, MAX, or AVG.
SELECT CustomerID, SUM(Price)
FROM Orders2
GROUP BY CustomerID
ORDER BY SUM(Price) DESC;
-- more complex example of GROUP BY
SELECT T.[Group] AS N'Region',
T.CountryRegionCode AS N'Country',
S.Name AS N'Store',
H.SalesPersonID,
SUM(TotalDue) AS N'Total Sales'
FROM Sales.Customer C
INNER JOIN Sales.Store S
ON C.StoreID = S.BusinessEntityID
INNER JOIN Sales.SalesTerritory T
ON C.TerritoryID = T.TerritoryID
INNER JOIN Sales.SalesOrderHeader H
ON C.CustomerID = H.CustomerID
WHERE T.[Group] = N'Europe'
AND T.CountryRegionCode IN(N'DE', N'FR')
AND H.SalesPersonID IN(287, 290, 288)
AND SUBSTRING(S.Name,1,4) IN(N'Vers', N'Spa ')
-- GROUP BY the columns in this order
GROUP BY T.[Group], T.CountryRegionCode, S.Name, H.SalesPersonID
ORDER BY T.[Group], T.CountryRegionCode, S.Name, H.SalesPersonID;
/* Results from GROUP BY above:
Region Country Store SalesPersonID Total Sales
-------- ------- --------------------------------- ------------- -------------
Europe DE Versatile Sporting Goods Company 287 729.6344
Europe DE Versatile Sporting Goods Company 288 17073.0655
Europe FR Spa and Exercise Outfitters 287 27731.551
Europe FR Spa and Exercise Outfitters 290 208479.3505
*/
-------------------
-- GROUPING SETS --
-------------------
/*
GROUPING SETS - an extention of GROUP BY that allows multiple grouping, good for things like generating grand total. When a NULL comes up in one of the cells, that NULL could be there because it's grouping multiple values (though it might all just be a NULL that's present in your data). GROUPING SETS is how you can identify how you want to group everything, where there are some shortcut groupings that do common grouping tasks like ROLLUP (shows grouping for a heirarchy) and CUBE (which groups by all possible grouping combinations).
SELECT [column list with aggregate(s)
FROM [source]
GROUP BY
GROUPING SETS
(
[columnname], -- one or more columns
[columnname], -- one or more columns
() -- empty parentheses if aggregating all rows
);
*/
--example of GROUP BY GROUPING SETS
SELECT T.[Group] AS N'Region',
T.CountryRegionCode AS N'Country',
S.Name AS N'Store',
H.SalesPersonID,
SUM(TotalDue) AS N'Total Sales'
FROM Sales.Customer C
INNER JOIN Sales.Store S
ON C.StoreID = S.BusinessEntityID
INNER JOIN Sales.SalesTerritory T
ON C.TerritoryID = T.TerritoryID
INNER JOIN Sales.SalesOrderHeader H
ON C.CustomerID = H.CustomerID
WHERE T.[Group] = N'Europe'
AND T.CountryRegionCode IN(N'DE', N'FR')
AND H.SalesPersonID IN(287, 290, 288)
AND SUBSTRING(S.Name,1,4) IN(N'Vers', N'Spa ')
--GROUP BY GROUPING SETS ending with () which will give a grand total
GROUP BY GROUPING SETS (T.[Group], T.CountryRegionCode, S.Name, H.SalesPersonID, ())
ORDER BY T.[Group], T.CountryRegionCode, S.Name, H.SalesPersonID;
/* Results from GROUP BY GROUPING SETS above. The first line with all NULLs is a grand total. The second line with all NULLs except SalesPersonID is the total for salesperson287. The bottom line with all NULLs except for Region is the total for Europe. etc.
Region Country Store SalesPersonID Total Sales
------- ------- --------------------------------- ------------- ---------------
NULL NULL NULL NULL 254013.6014
NULL NULL NULL 287 28461.1854
NULL NULL NULL 288 17073.0655
NULL NULL NULL 290 208479.3505
NULL NULL Spa and Exercise Outfitters NULL 236210.9015
NULL NULL Versatile Sporting Goods Company NULL 17802.6999
NULL DE NULL NULL 17802.6999
NULL FR NULL NULL 236210.9015
Europe NULL NULL NULL 254013.6014
*/
---------------------
-- GROUP BY ROLLUP --
---------------------
-- GROUP BY ROLLUP - provides a shortcut for defining groups sets with combinations that assume input columns from a hierarchy. Will give grand totals and subtotals for combinations.
--example of GROUP BY ROLLUP
SELECT T.[Group] AS N'Region',
T.CountryRegionCode AS N'Country',
S.Name AS N'Store',
H.SalesPersonID,
SUM(TotalDue) AS N'Total Sales'
FROM Sales.Customer C
INNER JOIN Sales.Store S
ON C.StoreID = S.BusinessEntityID
INNER JOIN Sales.SalesTerritory T
ON C.TerritoryID = T.TerritoryID
INNER JOIN Sales.SalesOrderHeader H
ON C.CustomerID = H.CustomerID
WHERE T.[Group] = N'Europe'
AND T.CountryRegionCode IN(N'DE', N'FR')
AND H.SalesPersonID IN(287, 290, 288)
AND SUBSTRING(S.Name,1,4) IN(N'Vers', N'Spa ')
--GROUP BY ROLLUP, showing subtotals by hierarchy
GROUP BY ROLLUP (T.[Group], T.CountryRegionCode, S.Name, H.SalesPersonID)
ORDER BY T.[Group], T.CountryRegionCode, S.Name, H.SalesPersonID;
/* Results from GROUP BY ROLLUP above. Line one with all NULLs is a grand total. Line 2 with all NULLs but Europe is a grand total for Europe, etc.
Region Country Store SalesPersonID Total Sales
------- ------- --------------------------------- ------------- ---------------------
NULL NULL NULL NULL 254013.6014
Europe NULL NULL NULL 254013.6014
Europe DE NULL NULL 17802.6999
Europe DE Versatile Sporting Goods Company NULL 17802.6999
Europe DE Versatile Sporting Goods Company 287 729.6344
Europe DE Versatile Sporting Goods Company 288 17073.0655
Europe FR NULL NULL 236210.9015
Europe FR Spa and Exercise Outfitters NULL 236210.9015
Europe FR Spa and Exercise Outfitters 287 27731.551
Europe FR Spa and Exercise Outfitters 290 208479.3505
*/
-------------------
-- GROUP BY CUBE --
-------------------
-- GROUP BY CUBE - like ROLLUP, but shows every possible combination. Provides a shortcut for defining goruping sets in which all possible combinations of grouping sets created. This can be good if you're not sure which grouping you want to use, since it'll show you everything (scortched earth grouping).
--example of GROUP BY CUBE
SELECT T.[Group] AS N'Region',
T.CountryRegionCode AS N'Country',
S.Name AS N'Store',
H.SalesPersonID,
SUM(TotalDue) AS N'Total Sales'
FROM Sales.Customer C
INNER JOIN Sales.Store S
ON C.StoreID = S.BusinessEntityID
INNER JOIN Sales.SalesTerritory T
ON C.TerritoryID = T.TerritoryID
INNER JOIN Sales.SalesOrderHeader H
ON C.CustomerID = H.CustomerID
WHERE T.[Group] = N'Europe'
AND T.CountryRegionCode IN(N'DE', N'FR')
AND H.SalesPersonID IN(287, 290, 288)
AND SUBSTRING(S.Name,1,4) IN(N'Vers', N'Spa ')
--GROUP BY CUBE, showing all possible grouping combinations
GROUP BY CUBE (T.[Group], T.CountryRegionCode, S.Name, H.SalesPersonID)
ORDER BY T.[Group], T.CountryRegionCode, S.Name, H.SalesPersonID;
/* Results from GROUP BY CUBE:
Region Country Store SalesPersonID Total Sales
------- ------- --------------------------------- ------------- ---------------------
NULL NULL NULL NULL 254013.6014
NULL NULL NULL 287 28461.1854
NULL NULL NULL 288 17073.0655
NULL NULL NULL 290 208479.3505
NULL NULL Spa and Exercise Outfitters NULL 236210.9015
NULL NULL Spa and Exercise Outfitters 287 27731.551
NULL NULL Spa and Exercise Outfitters 290 208479.3505
NULL NULL Versatile Sporting Goods Company NULL 17802.6999
NULL NULL Versatile Sporting Goods Company 287 729.6344
NULL NULL Versatile Sporting Goods Company 288 17073.0655
NULL DE NULL NULL 17802.6999
NULL DE NULL 287 729.6344
NULL DE NULL 288 17073.0655
NULL DE Versatile Sporting Goods Company NULL 17802.6999
NULL DE Versatile Sporting Goods Company 287 729.6344
NULL DE Versatile Sporting Goods Company 288 17073.0655
NULL FR NULL NULL 236210.9015
NULL FR NULL 287 27731.551
NULL FR NULL 290 208479.3505
NULL FR Spa and Exercise Outfitters NULL 236210.9015
NULL FR Spa and Exercise Outfitters 287 27731.551
NULL FR Spa and Exercise Outfitters 290 208479.3505
Europe NULL NULL NULL 254013.6014
Europe NULL NULL 287 28461.1854
Europe NULL NULL 288 17073.0655
Europe NULL NULL 290 208479.3505
Europe NULL Spa and Exercise Outfitters NULL 236210.9015
Europe NULL Spa and Exercise Outfitters 287 27731.551
Europe NULL Spa and Exercise Outfitters 290 208479.3505
Europe NULL Versatile Sporting Goods Company NULL 17802.6999
Europe NULL Versatile Sporting Goods Company 287 729.6344
Europe NULL Versatile Sporting Goods Company 288 17073.0655
Europe DE NULL NULL 17802.6999
Europe DE NULL 287 729.6344
Europe DE NULL 288 17073.0655
Europe DE Versatile Sporting Goods Company NULL 17802.6999
Europe DE Versatile Sporting Goods Company 287 729.6344
Europe DE Versatile Sporting Goods Company 288 17073.0655
Europe FR NULL NULL 236210.9015
Europe FR NULL 287 27731.551
Europe FR NULL 290 208479.3505
Europe FR Spa and Exercise Outfitters NULL 236210.9015
Europe FR Spa and Exercise Outfitters 287 27731.551
Europe FR Spa and Exercise Outfitters 290 208479.3505
*/
-----------------
-- GROUPING_ID --
-----------------
-- GROUPING SETS, ROLLUP, and CUBE will return a NULL value in some columns to show that this is a summary row (for example, if it's showing totals for the Category column, the Category will be listed in that row but all other columns will show NULL, showing you that this is the summary for that Category). This NULL can be indistinguishable from a NULL in the data set. Therefore, the GROUPING_ID function tells you if the NULL is due to a summary row or if there's actually a NULL in the data.
--example of GROUPING_ID
SELECT T.[Group] AS N'Region',
T.CountryRegionCode AS N'Country',
S.Name AS N'Store',
--GROUPING_ID function, will return a 1 if a NULL in SalesPerson column below is result of grouping SUMMARY
GROUPING_ID(H.SalesPersonID) AS N'SalesPerson G_ID',
H.SalesPersonID AS N'SalesPerson',
SUM(TotalDue) AS N'Total Sales'
FROM Sales.Customer C
INNER JOIN Sales.Store S
ON C.StoreID = S.BusinessEntityID
INNER JOIN Sales.SalesTerritory T
ON C.TerritoryID = T.TerritoryID
INNER JOIN Sales.SalesOrderHeader H
ON C.CustomerID = H.CustomerID
WHERE T.[Group] = N'Europe'
AND T.CountryRegionCode IN(N'DE', N'FR')
AND H.SalesPersonID IN(287, 290, 288)
AND SUBSTRING(S.Name,1,4) IN(N'Vers', N'Spa ')
GROUP BY ROLLUP (T.[Group], T.CountryRegionCode, S.Name, H.SalesPersonID)
ORDER BY T.[Group], T.CountryRegionCode, S.Name, H.SalesPersonID;
/* Results from GROUPING_ID above. 'SalesPerson G_ID' column shows a 1 if SalesPerson column is grouped. In top line, SalesPerson is NULL, but SalesPerson G_ID is 1, showing that NULL is the result of grouping. In bottom line, SalesPerson is NULL, but SalesPerson G_ID is 0, showing that the NULL is not the result of grouping, and therefore that NULL is in the data itself.
Region Country Store SalesPerson G_ID SalesPerson Total Sales
------- ------- --------------------------------- ---------------- ----------- ---------------------
NULL NULL NULL 1 NULL 254013.6014
Europe NULL NULL 1 NULL 254013.6014
Europe DE NULL 1 NULL 17802.6999
Europe DE Versatile Sporting Goods Company 1 NULL 17802.6999
Europe DE Versatile Sporting Goods Company 0 287 729.6344
Europe DE Versatile Sporting Goods Company 0 288 17073.0655
Europe FR NULL 1 NULL 236210.9015
Europe FR Spa and Exercise Outfitters 1 NULL 236210.9015
Europe FR Spa and Exercise Outfitters 0 287 27731.551
Europe FR Spa and Exercise Outfitters 0 NULL 208479.3505
*/以上是关于sql GROUP BY，GROUPING SETS，ROLLUP，CUBE，GROUPING_ID的主要内容，如果未能解决你的问题，请参考以下文章
group by ，grouping sets ，cube ，rollup区别
group by ，grouping sets ，cube ，rollup区别
group by ，grouping sets ，cube ，rollup区别
从多个表中创建 Sql Server VIEW GROUPing BY，选择子查询作为别名
SQL - 多 SELECT 和 GROUP BY 以获取表的 MIN(value)
ORDER BY clause is not in GROUP BY.this is incompatible with sql_mode=only_full_group_by 异常分析解决