6. optimization - tu braunschweig
TRANSCRIPT
• Exercise 1.1+1.2
– Linearization
–Why is order important?
6. Optimization
Data Warehousing & OLAP – Wolf-Tilo Balke – Institut für Informationssysteme – TU Braunschweig 1
11
36 47
22 53
78 89
64
911
1116 1217
1012 1313
1518 1619
1414
175
1910
2515
2720
Jan (1)
Feb(2)
Feb(3)
Feb(4)
D1
D2
• Exercise 2. a
6. Optimization
R1
R3
R5
08 Qtr3
08 Qtr4
09 Qtr1
Time
R
R1
R3
R508 Qtr4
09 Qtr1
Time
R
Data Warehousing & OLAP – Wolf-Tilo Balke – Institut für Informationssysteme – TU Braunschweig 2
R4
R6R2
08 Qtr1
08 Qtr2
08 Qtr3
Locationa b c d e f g
R6R2
08 Qtr1
08 Qtr2
08 Qtr3
a b c d e f g
R41 R42
• Exercise 2. a
6. Optimization
R1
R3
R508 Qtr4
09 Qtr1
Time
R
R1
R3
R508 Qtr4
09 Qtr1
Time
R
Data Warehousing & OLAP – Wolf-Tilo Balke – Institut für Informationssysteme – TU Braunschweig 3
R6R2
08 Qtr1
08 Qtr2
08 Qtr3
a b c d e f g
R6R2
08 Qtr1
08 Qtr2
08 Qtr3
a b c d e f g
R41 R42R41 R42
• Exercise 2. a
6. Optimization
R1
R31
R508 Qtr4
09 Qtr1
Time
R
R1
R3
R508 Qtr4
09 Qtr1
Time
R
R32
Data Warehousing & OLAP – Wolf-Tilo Balke – Institut für Informationssysteme – TU Braunschweig 4
R6R2
08 Qtr1
08 Qtr2
08 Qtr3
a b c d e f g
R6R2
08 Qtr1
08 Qtr2
08 Qtr3
a b c d e f g
R41 R42 R41 R42
• Exercise 2. a
6. Optimization
R1 R508 Qtr4
09 Qtr1
Time
R
DCR31
R5
08 Qtr3
08 Qtr4
09 Qtr1
Time
R
R32 R12
R11
Data Warehousing & OLAP – Wolf-Tilo Balke – Institut für Informationssysteme – TU Braunschweig 5
R6R2
08 Qtr1
08 Qtr2
08 Qtr3
a b c d e f g
BA
On X, highest minimum rectangles are B and D = ‘b’, and lowest maximum are A and C = ‘a’
On Y, highest minimum rectangle is D = ’09Qtr1’, and lowest maximum is A = ’08Qtr2’
Dx = 1/3; Dy = 3/5; => D and A will create the new split nodes
R6R2
08 Qtr1
08 Qtr2
08 Qtr3
a b c d e f gR41
R42
• Exercise 2. b
6. Optimization
R11 R2
R
R12
Data Warehousing & OLAP – Wolf-Tilo Balke – Institut für Informationssysteme – TU Braunschweig 6
R31 R41
R31.1 R31.2 R41.1 R41.2 R5.3 R6.3R6.2
R32 R42
R32.1 R32.2 R42.1 R42.2 R42.3
R5 R6
R5.2R5.1 R6.1
• Exercise 2. c
6. Optimization
R31
08 Qtr4
09 Qtr1
Time
R
R32 R12
R11
Data Warehousing & OLAP – Wolf-Tilo Balke – Institut für Informationssysteme – TU Braunschweig 7
R5
R6R2
08 Qtr1
08 Qtr2
08 Qtr3
08 Qtr4
a b c d e f gR41
R42
• Exercise 2. c
6. Optimization
R11 R2
R
R12
Data Warehousing & OLAP – Wolf-Tilo Balke – Institut für Informationssysteme – TU Braunschweig 8
R31 R41
R31.1 R31.2 R41.1 R41.2 R5.3 R6.3R6.2
R32 R42
R32.1 R32.2 R42.1 R42.2 R42.3
R5 R6
R5.2R5.1 R6.1
• Exercise 3.1
– UB-Trees
6. Optimization
Data Warehousing & OLAP – Wolf-Tilo Balke – Institut für Informationssysteme – TU Braunschweig 9
Data part
8 178 17 39 5139 51
2828
• Exercise 3.1
–Multidimensional Hierarchical Clustering (MHC)
6. Optimization
Sector Brown Goods White Goods
ALL
0 1Sector Brown Goods White GoodsBrown Goods White Goods
ALL
0 1
Data Warehousing & OLAP – Wolf-Tilo Balke – Institut für Informationssysteme – TU Braunschweig 10
...
Item
ProductGroup
Category VideoAudio
Camcorder VCR
TR-780 TRV-30 GR-AX 200 GV-500 SLV-E800
...
...
ID 2 11 5 8 21
00 01
0 1
Surrogate 00100000
0000 0001 0000 0001 0010
00100001 00110000 00110001 00110010
32 33 48 49 50
...
Item
ProductGroup
Category VideoAudio VideoAudio
Camcorder VCR
TR-780 TRV-30TR-780 TRV-30 GR-AX 200 GV-500 SLV-E800
...
...
ID 2 11 5 8 21
00 01
0 1
Surrogate 00100000
0000 0001 0000 0001 0010
00100001 00110000 00110001 00110010
32 33 48 49 50
• Exercise 3.2
– Multi-component BI:
• x = 4 *y+z, where 0 ≤ y ≤2, and 0 ≤z ≤3, called <3,4>
basis encoding
6. Optimization
Month Dec Nov Oct Sep Aug Jul Jun Mai Apr Mar Feb Jan
x = 4 *y+z, where 0 ≤ y ≤2, and 0 ≤z ≤3, called <3,4>
basis encoding
• 5 = 4*1+1
Data Warehousing & OLAP – Wolf-Tilo Balke – Institut für Informationssysteme – TU Braunschweig 11
Month Dec Nov Oct Sep Aug Jul Jun Mai Apr Mar Feb Jan
M A11 A10 A9 A8 A7 A6 A5 A4 A3 A2 A1 A0
5 0 0 0 0 0 0 1 0 0 0 0 0
X Y Z
M A2,1 A1,1 A0,1 A3,0 A2,0 A1,0 A0,0
5 0 1 0 0 0 1 0
• Exercise 3.2
– Range encoded BI
6. Optimization
Month Dec Nov Oct Sep Aug Jul Jun Mai Apr Mar Feb Jan
M A11 A10 A9 A8 A7 A6 A5 A4 A3 A2 A1 A0
0 1 1 1 1 1 1 1 1 1 1 1 1
– Persons born in March: ((NOT A1) AND A2)
Data Warehousing & OLAP – Wolf-Tilo Balke – Institut für Informationssysteme – TU Braunschweig 12
0 1 1 1 1 1 1 1 1 1 1 1 1
3 1 1 1 1 1 1 1 1 1 0 0 0
5 1 1 1 1 1 1 1 0 0 0 0 0
11 1 0 0 0 0 0 0 0 0 0 0 0