Related
I have following table
SELECT TableCode, Col1, Col2
FROM TableA
WHERE TableCode = 23
Result of Table:
TableCode | Col1 | Col1
23 | CustCode | QS
23 | CatCode | QS
After that i wrote one query on TableA which return following output
Query :
SELECT TableCode,x.ColCode,
x.ColumnName + '_' + CONVERT(VARCHAR(5), ROW_NUMBER() OVER (PARTITION BY X.COL ORDER BY X.COL)) [ColumnName],X.Values,
ROW_NUMBER() OVER (PARTITION BY X.COL ORDER BY X.COL) [RowNo]
FROM TableA a CROSS APPLY
(SELECT 1 ColCode,'ParaName' ColumnName,Col1 Values
UNION ALL
SELECT 2,'ParaSource',Col2
) x
WHERE TableCode = 23;
Result :
TableCode | ColCode | ColumnName | Values | RowNo
23 | 1 | ParaName_1 | CustCode | 1
23 | 1 | ParaName_2 | CatCode | 2
23 | 2 | ParaSource_1 | QS | 1
23 | 2 | ParaSource_2 | QS | 2
And i required following output:
Required Output :
TableCode | ColCode | ColumnName | Values | RowNo
23 | 1 | ParaName_1 | CustCode | 1
23 | 1 | ParaName_2 | CatCode | 2
23 | 1 | ParaName_3 | Null | 3
23 | 2 | ParaSource_1 | QS | 1
23 | 2 | ParaSource_2 | QS | 2
23 | 2 | ParaSource_3 | null | 3
Using a couple of common table expressions and row_number() along with the table value constructor (values (...),(...))
to cross join numbers 1, 2, and 3 then using a left join to return 3 rows per TableCode even when you do not have three rows in the source table.
;with numbered as (
select *, rn = row_number() over (order by (select 1))
from TableA
where TableCode = 23
)
, cte as (
select distinct tc.TableCode, a.Col1, a.Col2, v.rn
from numbered tc
cross join (values (1),(2),(3)) v (rn)
left join numbered a
on a.TableCode = tc.TableCode
and a.rn = v.rn
)
select
a.TableCode
, x.ColCode
, [ColumnName] = x.ColumnName + '_' + convert(varchar(5),a.rn)
, X.Value
,[RowNo] = a.rn
from cte a
cross apply (values (1,'ParaName',Col1),(2,'ParaSource',Col2))
as x(ColCode, ColumnName, Value)
order by ColCode, RowNo;
rextester demo: http://rextester.com/CJU8986
returns:
+-----------+---------+--------------+----------+-------+
| TableCode | ColCode | ColumnName | Value | RowNo |
+-----------+---------+--------------+----------+-------+
| 23 | 1 | ParaName_1 | CustCode | 1 |
| 23 | 1 | ParaName_2 | CatCode | 2 |
| 23 | 1 | ParaName_3 | NULL | 3 |
| 23 | 2 | ParaSource_1 | QS | 1 |
| 23 | 2 | ParaSource_2 | QS | 2 |
| 23 | 2 | ParaSource_3 | NULL | 3 |
+-----------+---------+--------------+----------+-------+
This would appear to do what you want:
SELECT TableCode, x.ColCode, v.*
FROM TableA a CROSS APPLY
(VALUES (1, 'ParaName-1', Col1, 1),
(2, 'ParaName-2', Col2, 2),
(3, 'ParaName-3', NULL, 2)
) v(ColCode, ColumnName, [Values], RowNo)
WHERE TableCode = 23;
I see no reason to use row_number() when you can just read in the correct values. Also, VALUES is a SQL keyword so it is a really bad column name.
My table has three rows:
ID MULTI_CODES
1 10-101-102
2 20-201-202
3 30-301-302
How to write a statement to split each row like this considering delimited code
ID SINGLE_CODE LEVEL
1 10 1
1 101 2
1 102 3
2 20 1
2 201 2
2 202 3
3 30 1
3 301 2
3 302 3
If you are on SQL Server 2016 you can use the built in string_split function.
If not you will need your own function. One that performs very well is Jeff Moden's tally table method. My altered version of this looks like this, allowing for user specified delimiters and if required specific values:
create function dbo.StringSplit
(
#str nvarchar(4000) = ' ' -- String to split.
,#delimiter as nvarchar(1) = ',' -- Delimiting value to split on.
,#num as int = null -- Which value to return.
)
returns table
as
return
(
-- Start tally table with 10 rows.
with n(n) as (select n from (values(1),(1),(1),(1),(1),(1),(1),(1),(1),(1)) n(n))
-- Select the same number of rows as characters in isnull(#str,'') as incremental row numbers.
-- Cross joins increase exponentially to a max possible 10,000 rows to cover largest isnull(#str,'') length.
,t(t) as (select top (select len(isnull(#str,'')) a) row_number() over (order by (select null)) from n n1,n n2,n n3,n n4)
-- Return the position of every value that follows the specified delimiter.
,s(s) as (select 1 union all select t+1 from t where substring(isnull(#str,''),t,1) = #delimiter)
-- Return the start and length of every value, to use in the SUBSTRING function.
-- ISNULL/NULLIF combo handles the last value where there is no delimiter at the end of the string.
,l(s,l) as (select s,isnull(nullif(charindex(#delimiter,isnull(#str,''),s),0)-s,4000) from s)
select rn as ItemNumber
,Item
from(select row_number() over(order by s) as rn
,substring(isnull(#str,''),s,l) as item
from l
) a
where rn = #num -- Return a specific value where specified,
or #num is null -- Or everything where not.
)
go
And is used as follows:
declare #t table (ID int, MULTI_CODES nvarchar(50));
insert into #t values (1,'10-101-102'),(2,'20-201-202'),(3,'30-301-302');
select t.ID
,t.MULTI_CODES
,s.Item as SINGLE_CODE
,s.ItemNumber as [Level]
from #t t
outer apply dbo.StringSplit(t.MULTI_CODES,'-',null) s
order by t.ID
,s.ItemNumber;
Which outputs:
+----+-------------+-------------+-------+
| ID | MULTI_CODES | SINGLE_CODE | Level |
+----+-------------+-------------+-------+
| 1 | 10-101-102 | 10 | 1 |
| 1 | 10-101-102 | 101 | 2 |
| 1 | 10-101-102 | 102 | 3 |
| 2 | 20-201-202 | 20 | 1 |
| 2 | 20-201-202 | 201 | 2 |
| 2 | 20-201-202 | 202 | 3 |
| 3 | 30-301-302 | 30 | 1 |
| 3 | 30-301-302 | 301 | 2 |
| 3 | 30-301-302 | 302 | 3 |
+----+-------------+-------------+-------+
IF OBJECT_ID('tempdb..#TEMPtable') IS NOT NULL
Drop table #TEMPtable
;With cte(ID, MULTI_CODES)
AS
(
select 1,'10-101-102' UNION ALL
select 2,'20-201-202' UNION ALL
select 3,'30-301-302'
)
SELECT * INTO #TEMPtable FROM cte
SELECT ID, Split.a.value('.','Varchar(100)') AS MULTI_CODES,ROW_NUMBER()Over(Partition by ID Order by ID) AS LEVEL
FROM(
SELECT ID, CASt('<M>' + Replace(MULTI_CODES,'-','</M><M>') +'</M>' AS XML)As MULTI_CODES
FROM #TEMPtable
)AS A
CROSS APPLY
MULTI_CODES.nodes('/M') AS Split(A)
OutPut
ID MULTI_CODES LEVEL
----------------------
1 10 1
1 101 2
1 102 3
2 20 1
2 201 2
2 202 3
3 30 1
3 301 2
3 302 3
you can use cross apply to string_split function
select id, Value as Single_code, RowN as [Level] from #yourcodes cross apply
(
select RowN= row_number() over (order by (select null)), value
from string_split(Multi_codes, '-')
) a
Output:
+----+-------------+-------+
| id | Single_code | Level |
+----+-------------+-------+
| 1 | 10 | 1 |
| 1 | 101 | 2 |
| 1 | 102 | 3 |
| 2 | 20 | 1 |
| 2 | 201 | 2 |
| 2 | 202 | 3 |
| 3 | 30 | 1 |
| 3 | 301 | 2 |
| 3 | 302 | 3 |
+----+-------------+-------+
I don't understand how PIVOT works in SQL. I have 2 tables and I would like to pivot 1 of them in order to get only 1 table with all the data together. I've attached an image with the tables I have and the result that I would like to get.
CREATE TABLE TABLE1
([serie_id] varchar(4), [Maturity] int, [Strategy] int, [Lifetime] varchar(4), [L_max] decimal(10, 5), [W_max] decimal(10, 5), [H_max] decimal(10, 5))
;
INSERT INTO TABLE1
([serie_id], [Maturity], [Strategy], [Lifetime], [L_max], [W_max], [H_max])
VALUES
('id_1', 3, 1, '2', 2.200, 1.400, 1.400),
('id_2', 3, 1, '2', 3.400, 1.800, 2.100),
('id_3', 3, 1, NULL, 24.500, 14.500, 15.000),
('id_4', 3, 1, NULL, 28.000, 24.500, 14.000)
;
CREATE TABLE TABLE2
([serie_id] varchar(4), [L_value] decimal(10, 5), [lrms] decimal(10, 5), [latTmax] decimal(10, 5), [Rdc] decimal(10, 5))
;
INSERT INTO TABLE2
([serie_id], [L_value], [lrms], [latTmax], [Rdc])
VALUES
('id_1', 67.000, 400.000, 400.000, 0.250),
('id_1', 90.000, 330.000, 330.000, 0.350),
('id_1', 120.000, 370.000, 370.000, 0.300),
('id_1', 180.000, 330.000, 300.000, 0.350),
('id_2', 260.000, 300.000, 300.000, 0.400),
('id_2', 360.000, 280.000, 280.000, 0.450),
('id_3', 90.000, 370.000, 370.000, 0.300),
('id_4', 160.000, 340.000, 340.000, 0.400)
;
SQLFiddle
If someone could help me with the SQL query I would appreciate it so much.
In order to get your final result, you are going to have to implement a variety of methods including unpivot, pivot, along with the use of a windowing function like row_number().
Since you have multiple columns in Table2 that need to be pivoted, then you will need to unpivot them first. This is the reverse of pivot, which converts your multiple columns into multiple rows. But before you unpivot, you need some value to identify the values of each row using row_number() - sounds complicated, right?
First, query table2 using the windowing function row_number(). This creates a unique identifier for each row and allows you to easily be able to associate the values for id_1 from any of the others.
select serie_id, l_value, lrms, latTmax, Rdc,
rn = cast(row_number() over(partition by serie_id order by serie_id)
as varchar(10))
from table2;
See Demo. Once you've created this unique identifier, then you will unpivot the L_value, lrms, latTmax, and rdc. You can unpivot the data using several different methods, including the unpivot function, CROSS APPLY, or UNION ALL.
select serie_id,
col, value
from
(
select serie_id, l_value, lrms, latTmax, Rdc,
rn = cast(row_number() over(partition by serie_id order by serie_id)
as varchar(10))
from table2
) d
cross apply
(
select 'L_value_'+rn, L_value union all
select 'lrms_'+rn, lrms union all
select 'latTmax_'+rn, latTmax union all
select 'Rdc_'+rn, Rdc
) c (col, value)
See SQL Fiddle with Demo. The data from table2 is not in a completely different format that can be pivoted into the new columns:
| SERIE_ID | COL | VALUE |
|----------|-----------|-------|
| id_1 | L_value_1 | 67 |
| id_1 | lrms_1 | 400 |
| id_1 | latTmax_1 | 400 |
| id_1 | Rdc_1 | 0.25 |
| id_1 | L_value_2 | 90 |
| id_1 | lrms_2 | 330 |
| id_1 | latTmax_2 | 330 |
| id_1 | Rdc_2 | 0.35 |
The final step would be to PIVOT the data above into the final result:
select serie_id, maturity, strategy, lifetime, l_max, w_max, h_max,
L_value_1, lrms_1, latTmax_1, Rdc_1,
L_value_2, lrms_2, latTmax_2, Rdc_2,
L_value_3, lrms_3, latTmax_3, Rdc_3,
L_value_4, lrms_4, latTmax_4, Rdc_4
from
(
select t1.serie_id, t1.maturity, t1.strategy, t1.lifetime,
t1.l_max, t1.w_max, t1.h_max,
t2.col, t2.value
from table1 t1
inner join
(
select serie_id,
col, value
from
(
select serie_id, l_value, lrms, latTmax, Rdc,
rn = cast(row_number() over(partition by serie_id order by serie_id)
as varchar(10))
from table2
) d
cross apply
(
select 'L_value_'+rn, L_value union all
select 'lrms_'+rn, lrms union all
select 'latTmax_'+rn, latTmax union all
select 'Rdc_'+rn, Rdc
) c (col, value)
) t2
on t1.serie_id = t2.serie_id
) d
pivot
(
max(value)
for col in (L_value_1, lrms_1, latTmax_1, Rdc_1,
L_value_2, lrms_2, latTmax_2, Rdc_2,
L_value_3, lrms_3, latTmax_3, Rdc_3,
L_value_4, lrms_4, latTmax_4, Rdc_4)
) p;
See SQL Fiddle with Demo.
If you have an unknown number of values in Table2 then you will need to use dynamic SQL to create a sql string that will be executed. Converting the above code to dynamic sql is pretty easy once you have the logic correct. The code will be:
DECLARE #cols AS NVARCHAR(MAX),
#query AS NVARCHAR(MAX)
select #cols
= STUFF((SELECT ',' + QUOTENAME(col+cast(rn as varchar(10)))
from
(
select rn = cast(row_number() over(partition by serie_id order by serie_id)
as varchar(10))
from table2
) d
cross apply
(
select 'L_value_', 0 union all
select 'lrms_', 1 union all
select 'latTmax_', 2 union all
select 'Rdc_', 3
) c (col, so)
group by col, rn, so
order by rn, so
FOR XML PATH(''), TYPE
).value('.', 'NVARCHAR(MAX)')
,1,1,'')
set #query = N'SELECT serie_id, maturity, strategy, lifetime, l_max,
w_max, h_max,' + #cols + N'
from
(
select t1.serie_id, t1.maturity, t1.strategy, t1.lifetime,
t1.l_max, t1.w_max, t1.h_max,
t2.col, t2.value
from table1 t1
inner join
(
select serie_id,
col, value
from
(
select serie_id, l_value, lrms, latTmax, Rdc,
rn = cast(row_number() over(partition by serie_id order by serie_id)
as varchar(10))
from table2
) d
cross apply
(
select ''L_value_''+rn, L_value union all
select ''lrms_''+rn, lrms union all
select ''latTmax_''+rn, latTmax union all
select ''Rdc_''+rn, Rdc
) c (col, value)
) t2
on t1.serie_id = t2.serie_id
) x
pivot
(
max(value)
for col in (' + #cols + N')
) p '
exec sp_executesql #query
See SQL Fiddle with Demo
Both versions will give a result of:
| SERIE_ID | MATURITY | STRATEGY | LIFETIME | L_MAX | W_MAX | H_MAX | L_VALUE_1 | LRMS_1 | LATTMAX_1 | RDC_1 | L_VALUE_2 | LRMS_2 | LATTMAX_2 | RDC_2 | L_VALUE_3 | LRMS_3 | LATTMAX_3 | RDC_3 | L_VALUE_4 | LRMS_4 | LATTMAX_4 | RDC_4 |
|----------|----------|----------|----------|-------|-------|-------|-----------|--------|-----------|-------|-----------|--------|-----------|--------|-----------|--------|-----------|--------|-----------|--------|-----------|--------|
| id_1 | 3 | 1 | 2 | 2.2 | 1.4 | 1.4 | 67 | 400 | 400 | 0.25 | 90 | 330 | 330 | 0.35 | 120 | 370 | 370 | 0.3 | 180 | 330 | 300 | 0.35 |
| id_2 | 3 | 1 | 2 | 3.4 | 1.8 | 2.1 | 260 | 300 | 300 | 0.4 | 360 | 280 | 280 | 0.45 | (null) | (null) | (null) | (null) | (null) | (null) | (null) | (null) |
| id_3 | 3 | 1 | (null) | 24.5 | 14.5 | 15 | 90 | 370 | 370 | 0.3 | (null) | (null) | (null) | (null) | (null) | (null) | (null) | (null) | (null) | (null) | (null) | (null) |
| id_4 | 3 | 1 | (null) | 28 | 24.5 | 14 | 160 | 340 | 340 | 0.4 | (null) | (null) | (null) | (null) | (null) | (null) | (null) | (null) | (null) | (null) | (null) | (null) |
I have a table like this . Table 1.
Customer Product Value Quantity Order Number
Dave Product 1 15 1 154
Dave Product 2 25 5 154
Dave Product 3 45 4 15
Rob Product 2 222 33 233
Now I want this , Table 2.
Customer Product 1 Quantity Product 1 Value Price per item ( Value /Quantity) for Product 1 Product 2 Quantity Product 2 Value Price per item ( Value /Quantity) for Product 2 Product 3 Quantity Product 3 Value Price per item ( Value /Quantity) for Product 3 Order Number
Dave 1 15 15 5 25 5 null null null 154
Dave null null null null null Null 4 45 11.25 15
Rob null null null 33 222 6.727272727 null null null 233
I was thinking about some pivot but wasn't sure how to construct it . Also the number of products is not fixed and change in Table 1.
In order to get the result, I would advise applying both an unpivot and a pivot to the data.
The UNPIVOT will convert the column data from your table into rows. Once the data is unpivoted, then you can apply a pivot.
Since you are using SQL Server 2008+ you can use CROSS APPLY with the VALUES clause to unpivot. Prior to 2008, you could use the UNPIVOT function. The code to unpivot the data is:
select t.customer,
replace(t.product, ' ', '')+'_'+c.col piv_col,
c.val,
t.ordernumber
from table1 t
cross apply
(
values
('value', cast(value as varchar(10))),
('quantity', cast(quantity as varchar(10))),
('PricePerUnit', cast((value/quantity) *1.0 as varchar(10)))
) c (col, val);
See Demo. This converts the data into the following format:
| CUSTOMER | PIV_COL | VAL | ORDERNUMBER |
---------------------------------------------------------
| Dave | Product1_value | 15 | 154 |
| Dave | Product1_quantity | 1 | 154 |
| Dave | Product1_PricePerUnit | 15.0 | 154 |
| Dave | Product2_value | 25 | 154 |
You can see that the row for Dave order 154 has been turned into rows and I have created the new column names that will be used for the pivot (piv_col). This column has concatenated the Product Name to the from of the previous column headers (value, quantity).
Since the data is in a single row, you can easily apply the pivot function to the data. The final code will be:
select customer,
Product1_quantity, Product1_value, Product1_PricePerUnit,
Product2_quantity, Product2_value, Product2_PricePerUnit,
Product3_quantity, Product3_value, Product3_PricePerUnit,
orderNumber
from
(
select t.customer,
replace(t.product, ' ', '')+'_'+c.col piv_col,
c.val,
t.ordernumber
from table1 t
cross apply
(
values
('value', cast(value as varchar(10))),
('quantity', cast(quantity as varchar(10))),
('PricePerUnit', cast((value/quantity) *1.0 as varchar(10)))
) c (col, val)
) d
pivot
(
max(val)
for piv_col in(Product1_quantity, Product1_value, Product1_PricePerUnit,
Product2_quantity, Product2_value, Product2_PricePerUnit,
Product3_quantity, Product3_value, Product3_PricePerUnit)
) piv;
See SQL Fiddle with Demo.
The above works great if you have a known number of products, but if not, then you will need to use dynamic SQL.
DECLARE #cols AS NVARCHAR(MAX),
#query AS NVARCHAR(MAX)
select #cols = STUFF((SELECT ',' + QUOTENAME(replace(t.product, ' ', '')+'_'+c.col)
from Table1 t
cross apply
(
values ('value', 1), ('quantity', 0),('PricePerUnit', 3)
) c (col, so)
group by product, col, so
order by product, so
FOR XML PATH(''), TYPE
).value('.', 'NVARCHAR(MAX)')
,1,1,'')
set #query = 'SELECT customer, ' + #cols + ', ordernumber
from
(
select t.customer,
replace(t.product, '' '', '''')+''_''+c.col piv_col,
c.val,
t.ordernumber
from table1 t
cross apply
(
values
(''value'', cast(value as varchar(10))),
(''quantity'', cast(quantity as varchar(10))),
(''PricePerUnit'', cast((value/quantity) *1.0 as varchar(10)))
) c (col, val)
) d
pivot
(
max(val)
for piv_col in (' + #cols + ')
) p '
execute(#query);
See SQL Fiddle with Demo. These queries give the result:
| CUSTOMER | PRODUCT1_QUANTITY | PRODUCT1_VALUE | PRODUCT1_PRICEPERUNIT | PRODUCT2_QUANTITY | PRODUCT2_VALUE | PRODUCT2_PRICEPERUNIT | PRODUCT3_QUANTITY | PRODUCT3_VALUE | PRODUCT3_PRICEPERUNIT | ORDERNUMBER |
-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
| Dave | (null) | (null) | (null) | (null) | (null) | (null) | 4 | 45 | 11.0 | 15 |
| Dave | 1 | 15 | 15.0 | 5 | 25 | 5.0 | (null) | (null) | (null) | 154 |
| Rob | (null) | (null) | (null) | 33 | 222 | 6.0 | (null) | (null) | (null) | 233 |
I have been looking at CROSS / OUTER APPLY with a colleague and we're struggling to find real life examples of where to use them.
I've spent quite a lot of time looking at When should I use CROSS APPLY over INNER JOIN? and googling but the main (only) example seems pretty bizarre (using the rowcount from a table to determine how many rows to select from another table).
I thought this scenario may benefit from OUTER APPLY:
Contacts Table (contains 1 record for each contact)
Communication Entries Table (can contain a phone, fax, email for each contact)
But using subqueries, common table expressions, OUTER JOIN with RANK() and OUTER APPLY all seem to perform equally. I'm guessing this means the scenario isn't applicable to APPLY.
Please share some real life examples and help explain the feature!
Some uses for APPLY are...
1) Top N per group queries (can be more efficient for some cardinalities)
SELECT pr.name,
pa.name
FROM sys.procedures pr
OUTER APPLY (SELECT TOP 2 *
FROM sys.parameters pa
WHERE pa.object_id = pr.object_id
ORDER BY pr.name) pa
ORDER BY pr.name,
pa.name
2) Calling a Table Valued Function for each row in the outer query
SELECT *
FROM sys.dm_exec_query_stats AS qs
CROSS APPLY sys.dm_exec_query_plan(qs.plan_handle)
3) Reusing a column alias
SELECT number,
doubled_number,
doubled_number_plus_one
FROM master..spt_values
CROSS APPLY (SELECT 2 * CAST(number AS BIGINT)) CA1(doubled_number)
CROSS APPLY (SELECT doubled_number + 1) CA2(doubled_number_plus_one)
4) Unpivoting more than one group of columns
Assumes 1NF violating table structure....
CREATE TABLE T
(
Id INT PRIMARY KEY,
Foo1 INT, Foo2 INT, Foo3 INT,
Bar1 INT, Bar2 INT, Bar3 INT
);
Example using 2008+ VALUES syntax.
SELECT Id,
Foo,
Bar
FROM T
CROSS APPLY (VALUES(Foo1, Bar1),
(Foo2, Bar2),
(Foo3, Bar3)) V(Foo, Bar);
In 2005 UNION ALL can be used instead.
SELECT Id,
Foo,
Bar
FROM T
CROSS APPLY (SELECT Foo1, Bar1
UNION ALL
SELECT Foo2, Bar2
UNION ALL
SELECT Foo3, Bar3) V(Foo, Bar);
There are various situations where you cannot avoid CROSS APPLY or OUTER APPLY.
Consider you have two tables.
MASTER TABLE
x------x--------------------x
| Id | Name |
x------x--------------------x
| 1 | A |
| 2 | B |
| 3 | C |
x------x--------------------x
DETAILS TABLE
x------x--------------------x-------x
| Id | PERIOD | QTY |
x------x--------------------x-------x
| 1 | 2014-01-13 | 10 |
| 1 | 2014-01-11 | 15 |
| 1 | 2014-01-12 | 20 |
| 2 | 2014-01-06 | 30 |
| 2 | 2014-01-08 | 40 |
x------x--------------------x-------x
CROSS APPLY
There are many situation where we need to replace INNER JOIN with CROSS APPLY.
1. If we want to join 2 tables on TOP n results with INNER JOIN functionality
Consider if we need to select Id and Name from Master and last two dates for each Id from Details table.
SELECT M.ID,M.NAME,D.PERIOD,D.QTY
FROM MASTER M
INNER JOIN
(
SELECT TOP 2 ID, PERIOD,QTY
FROM DETAILS D
ORDER BY CAST(PERIOD AS DATE)DESC
)D
ON M.ID=D.ID
The above query generates the following result.
x------x---------x--------------x-------x
| Id | Name | PERIOD | QTY |
x------x---------x--------------x-------x
| 1 | A | 2014-01-13 | 10 |
| 1 | A | 2014-01-12 | 20 |
x------x---------x--------------x-------x
See, it generated results for last two dates with last two date's Id and then joined these records only in outer query on Id, which is wrong. To accomplish this, we need to use CROSS APPLY.
SELECT M.ID,M.NAME,D.PERIOD,D.QTY
FROM MASTER M
CROSS APPLY
(
SELECT TOP 2 ID, PERIOD,QTY
FROM DETAILS D
WHERE M.ID=D.ID
ORDER BY CAST(PERIOD AS DATE)DESC
)D
and forms he following result.
x------x---------x--------------x-------x
| Id | Name | PERIOD | QTY |
x------x---------x--------------x-------x
| 1 | A | 2014-01-13 | 10 |
| 1 | A | 2014-01-12 | 20 |
| 2 | B | 2014-01-08 | 40 |
| 2 | B | 2014-01-06 | 30 |
x------x---------x--------------x-------x
Here is the working. The query inside CROSS APPLY can reference the outer table, where INNER JOIN cannot do this(throws compile error). When finding the last two dates, joining is done inside CROSS APPLY ie, WHERE M.ID=D.ID.
2. When we need INNER JOIN functionality using functions.
CROSS APPLY can be used as a replacement with INNER JOIN when we need to get result from Master table and a function.
SELECT M.ID,M.NAME,C.PERIOD,C.QTY
FROM MASTER M
CROSS APPLY dbo.FnGetQty(M.ID) C
And here is the function
CREATE FUNCTION FnGetQty
(
#Id INT
)
RETURNS TABLE
AS
RETURN
(
SELECT ID,PERIOD,QTY
FROM DETAILS
WHERE ID=#Id
)
which generated the following result
x------x---------x--------------x-------x
| Id | Name | PERIOD | QTY |
x------x---------x--------------x-------x
| 1 | A | 2014-01-13 | 10 |
| 1 | A | 2014-01-11 | 15 |
| 1 | A | 2014-01-12 | 20 |
| 2 | B | 2014-01-06 | 30 |
| 2 | B | 2014-01-08 | 40 |
x------x---------x--------------x-------x
OUTER APPLY
1. If we want to join 2 tables on TOP n results with LEFT JOIN functionality
Consider if we need to select Id and Name from Master and last two dates for each Id from Details table.
SELECT M.ID,M.NAME,D.PERIOD,D.QTY
FROM MASTER M
LEFT JOIN
(
SELECT TOP 2 ID, PERIOD,QTY
FROM DETAILS D
ORDER BY CAST(PERIOD AS DATE)DESC
)D
ON M.ID=D.ID
which forms the following result
x------x---------x--------------x-------x
| Id | Name | PERIOD | QTY |
x------x---------x--------------x-------x
| 1 | A | 2014-01-13 | 10 |
| 1 | A | 2014-01-12 | 20 |
| 2 | B | NULL | NULL |
| 3 | C | NULL | NULL |
x------x---------x--------------x-------x
This will bring wrong results ie, it will bring only latest two dates data from Details table irrespective of Id even though we join with Id. So the proper solution is using OUTER APPLY.
SELECT M.ID,M.NAME,D.PERIOD,D.QTY
FROM MASTER M
OUTER APPLY
(
SELECT TOP 2 ID, PERIOD,QTY
FROM DETAILS D
WHERE M.ID=D.ID
ORDER BY CAST(PERIOD AS DATE)DESC
)D
which forms the following desired result
x------x---------x--------------x-------x
| Id | Name | PERIOD | QTY |
x------x---------x--------------x-------x
| 1 | A | 2014-01-13 | 10 |
| 1 | A | 2014-01-12 | 20 |
| 2 | B | 2014-01-08 | 40 |
| 2 | B | 2014-01-06 | 30 |
| 3 | C | NULL | NULL |
x------x---------x--------------x-------x
2. When we need LEFT JOIN functionality using functions.
OUTER APPLY can be used as a replacement with LEFT JOIN when we need to get result from Master table and a function.
SELECT M.ID,M.NAME,C.PERIOD,C.QTY
FROM MASTER M
OUTER APPLY dbo.FnGetQty(M.ID) C
And the function goes here.
CREATE FUNCTION FnGetQty
(
#Id INT
)
RETURNS TABLE
AS
RETURN
(
SELECT ID,PERIOD,QTY
FROM DETAILS
WHERE ID=#Id
)
which generated the following result
x------x---------x--------------x-------x
| Id | Name | PERIOD | QTY |
x------x---------x--------------x-------x
| 1 | A | 2014-01-13 | 10 |
| 1 | A | 2014-01-11 | 15 |
| 1 | A | 2014-01-12 | 20 |
| 2 | B | 2014-01-06 | 30 |
| 2 | B | 2014-01-08 | 40 |
| 3 | C | NULL | NULL |
x------x---------x--------------x-------x
Common feature of CROSS APPLY and OUTER APPLY
CROSS APPLY or OUTER APPLY can be used to retain NULL values when unpivoting, which are interchangeable.
Consider you have the below table
x------x-------------x--------------x
| Id | FROMDATE | TODATE |
x------x-------------x--------------x
| 1 | 2014-01-11 | 2014-01-13 |
| 1 | 2014-02-23 | 2014-02-27 |
| 2 | 2014-05-06 | 2014-05-30 |
| 3 | NULL | NULL |
x------x-------------x--------------x
When you use UNPIVOT to bring FROMDATE AND TODATE to one column, it will eliminate NULL values by default.
SELECT ID,DATES
FROM MYTABLE
UNPIVOT (DATES FOR COLS IN (FROMDATE,TODATE)) P
which generates the below result. Note that we have missed the record of Id number 3
x------x-------------x
| Id | DATES |
x------x-------------x
| 1 | 2014-01-11 |
| 1 | 2014-01-13 |
| 1 | 2014-02-23 |
| 1 | 2014-02-27 |
| 2 | 2014-05-06 |
| 2 | 2014-05-30 |
x------x-------------x
In such cases a CROSS APPLY or OUTER APPLY will be useful
SELECT DISTINCT ID,DATES
FROM MYTABLE
OUTER APPLY(VALUES (FROMDATE),(TODATE))
COLUMNNAMES(DATES)
which forms the following result and retains Id where its value is 3
x------x-------------x
| Id | DATES |
x------x-------------x
| 1 | 2014-01-11 |
| 1 | 2014-01-13 |
| 1 | 2014-02-23 |
| 1 | 2014-02-27 |
| 2 | 2014-05-06 |
| 2 | 2014-05-30 |
| 3 | NULL |
x------x-------------x
One real life example would be if you had a scheduler and wanted to see what the most recent log entry was for each scheduled task.
select t.taskName, lg.logResult, lg.lastUpdateDate
from task t
cross apply (select top 1 taskID, logResult, lastUpdateDate
from taskLog l
where l.taskID = t.taskID
order by lastUpdateDate desc) lg
To answer the point above knock up an example:
create table #task (taskID int identity primary key not null, taskName varchar(50) not null)
create table #log (taskID int not null, reportDate datetime not null, result varchar(50) not null, primary key(reportDate, taskId))
insert #task select 'Task 1'
insert #task select 'Task 2'
insert #task select 'Task 3'
insert #task select 'Task 4'
insert #task select 'Task 5'
insert #task select 'Task 6'
insert #log
select taskID, 39951 + number, 'Result text...'
from #task
cross join (
select top 1000 row_number() over (order by a.id) as number from syscolumns a cross join syscolumns b cross join syscolumns c) n
And now run the two queries with a execution plan.
select t.taskID, t.taskName, lg.reportDate, lg.result
from #task t
left join (select taskID, reportDate, result, rank() over (partition by taskID order by reportDate desc) rnk from #log) lg
on lg.taskID = t.taskID and lg.rnk = 1
select t.taskID, t.taskName, lg.reportDate, lg.result
from #task t
outer apply ( select top 1 l.*
from #log l
where l.taskID = t.taskID
order by reportDate desc) lg
You can see that the outer apply query is more efficient. (Couldn't attach the plan as I'm a new user... Doh.)