�ݺ�ߣ

Make your Data Dance
UNPIVOT, PIVOT and GROUP BY Extensions
Stew Ashton
UKOUG Techfest 2019
Can you read the following line? If not, please move closer.
It's much better when you can read the code ;)

Who am I?
• 1981-2015:
– Developer / application architect
– From Mainframe to client / server to Web
• 2005-present: focus on Oracle DB development
– Advocate of data-centric application architecture
– Contribute to asktom & ODC/OTN SQL forum
– Presented at OOW, conferences, meetups…
2

Data with Subtotals?
REGION COUNTRY CUSTOMER SALE_DATE QTY AMOUNT
Europe DE Raina Silverberg 2017-01-23 2 110
3
Europe DE Gloria Saintclair 2017-01-27 2 140
Europe DE Gloria Saintclair (Subtotal) 5 420
Europe DE Raina Silverberg (Subtotal) 5 330

Totals and Grand Total?
REGION COUNTRY CUSTOMER SALE_DATE QTY AMOUNT
4
Europe DE Raina Silverberg (Subtotal) 5 330
Europe DE Gloria Saintclair (Subtotal) 5 420
Europe DE (Subtotal) 10 750
Europe (Total) 10 750
(Grand total) 10 750

Columns to Rows?
5
A C
1A
2A
1
2
3A3
4
5
6
1B
2B
3B
7
8
9
1C
2C
3C
B
1B
2B
3B
1C
2C
3C

Rows to Columns?
6
A C
1A
2A
1
2
3A3
4
5
6
4A
5A
6A
7
8
9
7A
8A
9A
B
4A
5A
6A
7A
8A
9A

Transpose?
7
A C
1A
2A
1
2
3A3
B
1B
2B
3B
1C
2C
3C
1 32
A
B
C
Rows to columns
and
Columns to rows?

Rotate?
8
A C
1A
2A
1
2
3A3
B
1B
2B
3B
1C
2C
3C

How?
• UNPIVOT columns to rows
• PIVOT rows to columns
• GROUP BY extensions
– ROLLUP
– CUBE
– GROUPING SETS
• GROUPING_ID() and GROUPING() functions
9

UNPIVOT: columns to rows
10
KEY COL VAL
1 A 1A
KEY A B
1 1A 1B
1 B 1B
1
with data(KEY,A,B) as (
select 1, '1A', '1B' from dual
)
select * from data
unpivot(VAL for COL in(A,B));
KEY A B
1 1A 1B1
Metadata becomes data

KEYKEY 'A' 'B'
1A 1B1
PIVOT: rows to columns
11
KEY COL VAL
1 A 1A
1 B 1B
select * from data
unpivot(VAL for COL in(A,B));
pivot(max(VAL) for COL in('A','B')); Data becomes metadata
1 C 1C
Only listed COL values produce output

KEY A_N A_V
1 VA1
Multiple Columns
12
select * from data
unpivot((N, V) for COL in ((A_N, A_V) as 'A',(B_N, B_V) as 'B'))
pivot( max(N) as N, max(V) as V for COL in ('A' as A,'B' as B));
B_N B_V
2 VB
KEY COL N
A 11
V
VA
B 21 VB
1
Round trip!

KEY
1 VA1
PIVOT generates column names
13
pivot(
max(N) as N, max(V) as V
for COL in ('A' as A,'B' as B)
);
A_N A_V B_N B_V
2 VB1
A N A V B N B VA N A V B N B V

Transpose!
14
A C
1A
2A
1
2
3A3
B
1B
2B
3B
1C
2C
3C
1 32
A
B
C
select * from t
unpivot(x for nk in(a,b,c))
pivot(max(x) for k in(1,2,3)); KNK

Matrix: rotate!
15
A C
1A
2A
1
2
3A3
B
1B
2B
3B
1C
2C
3C
with normalized as (
select k r, c, v from t
unpivot(v for c in(
a as 1,b as 2,c as 3
))
)
, rotated as (
select 4-c r, r c, v
from normalized
)
select * from rotated
pivot(max(v) for c in(1,2,3))
order by r;
Normalize to cells (row, column, value),
rotate, display

Grouping Sales History
• 2 regions
• 2 countries per region
• 2 customers per country
• 2 years
• 2 sales per year per customer
16

17
select sum(amount) amount
from sales_yr;
AMOUNT
18360

18
select sum(amount) amount
from sales_yr
group by ();
AMOUNT
18360

19
select region,
sum(amount) amount
from sales_yr
group by region;
REGION AMOUNT
Americas 9720
Europe 8640
What if I want both?

20
select region,
sum(amount) amount
from sales_yr
group by grouping sets (
(),
region
);
REGION AMOUNT
Americas 9720
Europe 8640
18360

21
select region, country,
sum(amount) amount
from sales_yr
group by grouping sets (
(),
region,
(region, country)
);
REGION COUNTRY AMOUNT
Europe DE 1500
Europe FR 7140
Europe 8640
Americas CA 8820
Americas US 900
Americas 9720
18360
What if I want 3 levels?

22
sum(amount) amount
from sales_yr
group by rollup(region, country);
Europe DE 1500
Europe FR 7140
Europe 8640
Americas CA 8820
Americas US 900
Americas 9720
18360

23
sum(amount) amount
from sales_yr
group by cube(region, country);
18360
CA 8820
DE 1500
FR 7140
US 900
Europe 8640
Europe DE 1500
Europe FR 7140
Americas 9720
Americas CA 8820
Americas US 900
What if I have NULL data?

24
select
region, country,
sum(amount) amount
from sales_yr
18360
CA 8820
DE 1500
FR 7140
US 900
Europe 8640
Europe DE 1500
Europe FR 7140
Americas 9720
Americas CA 8820
Americas US 900

25
select
grouping(country) g_c,
region, country,
sum(amount) amount
from sales_yr
G_C REGION COUNTRY AMOUNT
1 18360
0 CA 8820
0 DE 1500
0 FR 7140
0 US 900
1 Europe 8640
0 Europe DE 1500
0 Europe FR 7140
1 Americas 9720
0 Americas CA 8820
0 Americas US 900

26
select
grouping(region)*2 g_r,
region, country,
sum(amount) amount
from sales_yr
G_R G_C REGION COUNTRY AMOUNT
2 1 18360
2 0 CA 8820
2 0 DE 1500
2 0 FR 7140
2 0 US 900
0 1 Europe 8640
0 0 Europe DE 1500
0 0 Europe FR 7140
0 1 Americas 9720
0 0 Americas CA 8820
0 0 Americas US 900

27
select
grouping_id(region, country) g_rc,
grouping(region)*2 g_r,
region, country,
sum(amount) amount
from sales_yr
G_RC G_R G_C REGION COUNTRY AMOUNT
3 2 1 18360
2 2 0 CA 8820
2 2 0 DE 1500
2 2 0 FR 7140
2 2 0 US 900
1 0 1 Europe 8640
0 0 0 Europe DE 1500
0 0 0 Europe FR 7140
1 0 1 Americas 9720
0 0 0 Americas CA 8820
0 0 0 Americas US 900

select * from (
select deptno, job, sal from emp
)
pivot(sum(sal) for deptno in(
10 as "10", 20 as "20", 30 as "30"
));
select deptno,
job, sum(sal) sal
from emp
group by deptno, job;
28
DEPTNO JOB SAL
10 CLERK 1300
10 MANAGER 2450
10PRESIDENT 5000
20 CLERK 1900
20 ANALYST 6000
20 MANAGER 2975
30 CLERK 950
30 MANAGER 2850
30SALESMAN 5600
JOB 10 20 30
ANALYST 6000
CLERK 1300 1900 950
MANAGER 2450 2975 2850
PRESIDENT 5000
SALESMAN 5600
SpreadSheet-like Data from EMP

select deptno,
job, sum(sal) sal
from emp
group by cube(deptno, job);
29
select
deptno,
job, sum(sal) sal
from emp
DEPTNO JOB SAL
29025
CLERK 4150
ANALYST 6000
MANAGER 8275
SALESMAN 5600
PRESIDENT 5000
10 8750
10 CLERK 1300
10 MANAGER 2450
10PRESIDENT 5000
20 10875
20 CLERK 1900
20 ANALYST 6000
20 MANAGER 2975
30 9400
30 CLERK 950
30 MANAGER 2850
30SALESMAN 5600
JOB 10 20 30 TOTAL
ANALYST 6000 6000
CLERK 1300 1900 950 4150
MANAGER 2450 2975 2850 8275
PRESIDENT 5000 5000
SALESMAN 5600 5600
(Total) 8750 10875 9400 29025
SpreadSheet-like Data from EMP

30
select case grouping(deptno)
when 1 then -1 else deptno
end deptno,
job, sum(sal) sal
from emp
DEPTNO JOB SAL
-1 29025
-1 CLERK 4150
-1 ANALYST 6000
-1 MANAGER 8275
-1SALESMAN 5600
-1PRESIDENT 5000
10 8750
10 CLERK 1300
10 MANAGER 2450
10PRESIDENT 5000
20 10875
20 CLERK 1900
20 ANALYST 6000
20 MANAGER 2975
30 9400
30 CLERK 950
30 MANAGER 2850
30SALESMAN 5600

31
end deptno,
job, sum(sal) sal,
grouping(job) gr_job
from emp
DEPTNO JOB SALGR_JOB
-1 29025 1
-1 CLERK 4150 0
-1 ANALYST 6000 0
-1 MANAGER 8275 0
-1SALESMAN 5600 0
-1PRESIDENT 5000 0
10 8750 1
10 CLERK 1300 0
10 MANAGER 2450 0
10PRESIDENT 5000 0
20 10875 1
20 CLERK 1900 0
20 ANALYST 6000 0
20 MANAGER 2975 0
30 9400 1
30 CLERK 950 0
30 MANAGER 2850 0
30SALESMAN 5600 0

32
select case gr_job
when 1 then '(Total)' else job
end job,
"10", "20", "30", "(Total)"
from (
end deptno,
job, sum(sal) sal,
grouping(job) gr_job
from emp
group by cube(deptno, job)
)
pivot(max(sal) for deptno in (
10, 20, 30, -1 as TOTAL
))
order by gr_job, job;
JOB 10 20 30 TOTAL
ANALYST 6000 6000
CLERK 1300 1900 950 4150
MANAGER 2450 2975 2850 8275
PRESIDENT 5000 5000
SALESMAN 5600 5600
(Total) 8750 10875 9400 29025

SpreadSheet-like Data from SALES
• 2 sums
– QTY, AMOUNT
• 4 vertical levels
– Grand total, region, country, customer
• 2 horizontal levels
– Year, total all years
33

34
select
sum(QTY) QTY,
sum(AMOUNT) AMOUNT
from sales_yr
group by grouping sets(
rollup(REGION,COUNTRY,CUSTOMER,YR),
rollup(YR,REGION,COUNTRY)
)

35
select
decode(grouping_id(REGION,COUNTRY,CUSTOMER),7,'Total',REGION) REGION,
decode(grouping_id(REGION,COUNTRY,CUSTOMER),3,'Total',COUNTRY) COUNTRY,
decode(grouping_id(REGION,COUNTRY,CUSTOMER),1,'Total',CUSTOMER) CUSTOMER,
decode(grouping(YR),1,'Total',YR) YR,
sum(QTY) QTY,
sum(AMOUNT) AMOUNT
from sales_yr
)

36
select * from (
select
decode(grouping_id(REGION,COUNTRY,CUSTOMER),7,'Total',REGION) REGION,
decode(grouping_id(REGION,COUNTRY,CUSTOMER),3,'Total',COUNTRY) COUNTRY,
decode(grouping_id(REGION,COUNTRY,CUSTOMER),1,'Total',CUSTOMER) CUSTOMER,
decode(grouping(YR),1,'Total',YR) YR,
sum(QTY) QTY,
sum(AMOUNT) AMOUNT
from sales_yr
)
)
pivot(max(qty) as q,max(amount) as a for (yr) in (
('2016') as "2016",
('2017') as "2017",
('Total') as "Total"
))
order by nullif(region,'Total'),nullif(country,'Total'),nullif(customer,'Total')

37
REGION COUNTRY CUSTOMER 2016_Q 2016_A 2017_Q 2017_A Total_Q Total_A
Americas CA Alyce Gray 13 1530 13 1530 26 3060
Americas CA Salena Grace 21 2880 21 2880 42 5760
Americas CA (Total) 34 4410 34 4410 68 8820
Americas US Linette Ingram 3 210 3 210 6 420
Americas US Vida Puleo 3 240 3 240 6 480
Americas US (Total) 6 450 6 450 12 900
Americas (Total) 40 4860 40 4860 80 9720
Europe DE Gloria Saintclair 5 420 5 420 10 840
Europe DE Raina Silverberg 5 330 5 330 10 660
Europe DE (Total) 10 750 10 750 20 1500
Europe FR Koren Macdonald 17 2100 17 2100 34 4200
Europe FR Madelaine Gottlieb 11 1470 11 1470 22 2940
Europe FR (Total) 28 3570 28 3570 56 7140
Europe (Total) 38 4320 38 4320 76 8640
(Total) 78 9180 78 9180 156 18360

Make your Data Dance
UNPIVOT, PIVOT and GROUP BY Extensions
Stew Ashton
UKOUG Techfest 2019
Blog: stewashton.wordpress.com
Twitter: @stewashton
/search/slideshow?&q=stewashton

UNPIVOT: Intersecting Ranges
39
C
1 (null)
2 5
3 4
# ranges 1 2 3 2 1
O
1 3
2 4
5 (null)
# ranges 1 2 1 1

40
OBJ_ID F T
C 1(null)
C 25
C 34
O 13
O 24
O 5(null)
select * from range_input

41
unpivot (
F for FT in (F as 1, T as -1)
)
OBJ_ID FT F
C 11
C 12
C -15
C 13
C -14
O 11
O -13
O 12
O -14
O 15
OBJ_ID F T
C 1(null)
C 25
C 34
O 13
O 24
O 5(null)

42
unpivot include nulls (
)
OBJ_ID FT F
C 11
C -1(null)
C 12
C -15
C 13
C -14
O 11
O -13
O 12
O -14
O 15
O -1(null)
OBJ_ID FT F
C 11
C 12
C -15
C 13
C -14
O 11
O -13
O 12
O -14
O 15

43
with unpivoted as (
unpivot include nulls (
)
)
select OBJ_ID, F, T, NUM_R
from (
select a.*,
lead(F) over(partition by obj_id order by F) T,
sum(FT) over(partition by obj_id order by F) NUM_R
from unpivoted a
) where F < nvl(T,F+1) and NUM_R > 0
order by obj_id, T;
OBJ_ID F T NUM_R
C 1 2 1
C 2 3 2
C 3 4 3
C 4 5 2
C 5 (null) 1
O 1 2 1
O 2 3 2
O 3 4 1
O 5 (null) 1

�ݺ�ߣ

Make your data dance: PIVOT, UNPIVOT & GROUP BY extensions

Recommended

More Related Content

Similar to Make your data dance: PIVOT, UNPIVOT & GROUP BY extensions (20)

Recently uploaded (20)

Make your data dance: PIVOT, UNPIVOT & GROUP BY extensions