Post on 17-Feb-2016
description
© 2009 Wipro Ltd – Internal & Restricted
Advanced OLAP ConceptsAdvanced OLAP ConceptsPart 1Part 1
<OLAP Concepts 201 Series>Renu Singh
© 2009 Wipro Ltd - Confidential2 © 2009 Wipro Ltd - Confidential2
OLAP Overview• On-Line Analytical Processing is a decision support
software that allows the user to quickly analyze information that has been summarized into multidimensional views and hierarchies.
• There are three main features of OLAP system : Multidimensional Viewing – OLAP supports
multidimensional model which consists of facts and dimensions also called as Star Schema.
Calculation Intensive Capabilities – Due to data is stored in facts and dimensions tables, it enables users to analyze data without much calculations.
Time Series analysis – Enables users to analyze data across time.
• This Module will cover the details in four parts:– Part 1 – OLAP Modeling – Building Blocks– Part 2 – OLAP Types & Architectures – Part 3 – OLAP Reporting Styles – Part 4 – OLAP Issues & Optimization
© 2009 Wipro Ltd - Confidential3 © 2009 Wipro Ltd - Confidential3
OLAP Objectives
Upon completion of this module you will be able to:• Understand OLAP Building Blocks i.e. OLAP Modeling
© 2009 Wipro Ltd - Confidential4 © 2009 Wipro Ltd - Confidential4
OLAP Outline
Lesson 1 OLAP Modeling
© 2009 Wipro Ltd - Confidential5
OLAP Modeling
© 2009 Wipro Ltd - Confidential6 © 2009 Wipro Ltd - Confidential6 © 2009 Wipro Ltd - Confidential6
• Multidimensional modeling– Basis for OLAP application. It represents data
under the metaphor of a cube whose cell corresponds to event that occurred in business domain.
Building Blocks - Dimensional Fact modeling
DimensionDimension
DimensionMeasure
© 2009 Wipro Ltd - Confidential7 © 2009 Wipro Ltd - Confidential7 © 2009 Wipro Ltd - Confidential7
Building Blocks - Dimensional Fact modeling
A simple representation of dimensional fact schema
Invoice Line
QuantityUnit priceNet amountVATTotal amountdiscount
yeardate month
week
month
Product
Type
brand
Category
state city
customer
© 2009 Wipro Ltd - Confidential8 © 2009 Wipro Ltd - Confidential8 © 2009 Wipro Ltd - Confidential8
Building Blocks - Dimensional Fact modeling
Dimension AttributesProperty of dimensionExample: product dimension is described by type, category, brand
HierarchyDirected tree, rooted in a dimension. All nodes are dimension attributesExample:
© 2009 Wipro Ltd - Confidential9 © 2009 Wipro Ltd - Confidential9 © 2009 Wipro Ltd - Confidential9
Building Blocks - Dimensional Fact modeling
Few other important items• Convergence - two attributes in a same dimension and same
hierarchy are connected by more than one alternate path
• Cross dimensional attribute – whose value is determined by two or more dimension attribute. Example: VAT- depends on both product category and state
• Operational/Multiple Arcs – Defines the association of attributes in dimensional model
© 2009 Wipro Ltd - Confidential10 © 2009 Wipro Ltd - Confidential10 © 2009 Wipro Ltd - Confidential10
• Shared Hierarchy – Helps to reduce redundancy of using large portion of hierarchies twice or more in same Fact schema
• Ragged (incomplete) Hierarchy – Relates the instances where values of one or more attributes are missing. It mostly occurs in geographic hierarchy
• Unbalanced Hierarchy – each level has a consistent meaning, but the branches have inconsistent depths
• Dynamic Hierarchy – Frequently changes hierarchy as time is a key factor
Building Blocks - Dimensional Fact modeling
© 2009 Wipro Ltd - Confidential11 © 2009 Wipro Ltd - Confidential11 © 2009 Wipro Ltd - Confidential11
OLAP Cube (Data Cube)It is a data structure that allows faster analysis of data. It will
help user to analyze facts at multiple level of abstraction
A cube is a set of possible views defined over a list of dimensions, a base table and aggregated measures. Cube view can be denoted as CV[G], where G is a granularity
Example: A cube view that sums amounts sold at category Department can be defined as ………
Select Department, SUM(amount)from sales, T[Product, department]where sales.product = T[Product, department].productgroup by department
This view can be defined as CV[department]
Building Blocks - Dimensional Fact modeling
© 2009 Wipro Ltd - Confidential12 © 2009 Wipro Ltd - Confidential12 © 2009 Wipro Ltd - Confidential12
OLAP Operations
Rollup and Drilldown operations:Rollup operations aggregates cube view to
a higher granularity. It is also called a summarization or consolidation. As an example, sales figure can be aggregated in agent level and then city level
Drilldown is a process where user travel from summarized to more detail level. As an example, sales in zonal level can be detailed to agent level
Building Blocks - Dimensional Fact modeling
© 2009 Wipro Ltd - Confidential13 © 2009 Wipro Ltd - Confidential13 © 2009 Wipro Ltd - Confidential13
OLAP Operations Slice and Dice operations:
Slice is limiting the analysis of OLAP cube to a given attribute/property. It is a two dimensional view of a cube.
Dice limits analysis to subset of an attribute
Building Blocks - Dimensional Fact modeling
© 2009 Wipro Ltd - Confidential14 © 2009 Wipro Ltd - Confidential14 © 2009 Wipro Ltd - Confidential14
Basic Features• Multidimensional conceptual view• Intuitive data manipulation• Accessibility• Client Server Architecture• Transparency• Multi-user support
Reporting Features• Flexible reporting• Uniform reporting performance• Automatic adjustment of Physical Level
Building Blocks – Codd Rules for OLAP
© 2009 Wipro Ltd - Confidential15 © 2009 Wipro Ltd - Confidential15 © 2009 Wipro Ltd - Confidential15
Dimensional Features• Generic dimensionality• Unlimited dimensions and aggregation levels• Unrestricted cross-dimensional operations
Building Blocks – Codd Rules for OLAP
© 2009 Wipro Ltd - Confidential16 © 2009 Wipro Ltd - Confidential16 © 2009 Wipro Ltd - Confidential16
Congratulations! You have now completed the module OLAP Building Blocks
You should now be able – Dimensional Modeling– Features of Dimensional Modeling– Codd Rules for OLAP
Module Summary
© 2009 Wipro Ltd – Internal & Restricted17
Quiz
© 2009 Wipro Ltd – Internal & Restricted18 © 2009 Wipro Ltd – Internal & Restricted18 © 2009 Wipro Ltd – Internal & Restricted18 © 2009 Wipro Ltd – Internal & Restricted18
1. Process to see low level of information from aggregated level isa) Drill Downb) Drill Upc) Drill Acrossd) Drill Through
Question 1
Answer: A
© 2009 Wipro Ltd – Internal & Restricted19 © 2009 Wipro Ltd – Internal & Restricted19 © 2009 Wipro Ltd – Internal & Restricted19 © 2009 Wipro Ltd – Internal & Restricted19
2. Lowest level of information stored in Dimensional model is called asa) Factb) Granularityc) Measured) None of the above
Question 2
Answer: B
© 2009 Wipro Ltd – Internal & Restricted20 © 2009 Wipro Ltd – Internal & Restricted20 © 2009 Wipro Ltd – Internal & Restricted20 © 2009 Wipro Ltd – Internal & Restricted20
3. Basis of OLAP system is <>a) Measureb) Factc) Dimensiond) Multi Dimensional Model
Question 3
Answer: D
© 2009 Wipro Ltd – Internal & Restricted21 © 2009 Wipro Ltd – Internal & Restricted21 © 2009 Wipro Ltd – Internal & Restricted21 © 2009 Wipro Ltd – Internal & Restricted21
4. _____ is a data structure that allows faster analysis of data. a) Measureb) OLAP Cubec) Dimensiond) None of the above
Question 4
Answer: B
© 2009 Wipro Ltd – Internal & Restricted22 © 2009 Wipro Ltd – Internal & Restricted22 © 2009 Wipro Ltd – Internal & Restricted22 © 2009 Wipro Ltd – Internal & Restricted22
5. Two attributes in a same dimension and same hierarchy are connected by more than one alternate patha) Convergenceb) Cross Dimensionc) Conformed Dimensiond) None of the above
Question 5
Answer: A
© 2009 Wipro Ltd – Internal & Restricted23
References
© 2009 Wipro Ltd – Internal & Restricted24 © 2009 Wipro Ltd – Internal & Restricted24 © 2009 Wipro Ltd – Internal & Restricted24 © 2009 Wipro Ltd – Internal & Restricted24
References
Library:Tek-tips reporting solutions forumhttp://www.dmreview.com/http://www.dwinfocenter.org
Data warehouse and OLAP – Concepts, Architecture and Solutions by Robert Wrember & Christian Koncilia
The OLAP Report by Nigel Pendse and Richard Creeth
The OLAP Solutions – Building Multidimensional Information Systems Second Edition by Erik Thomsen, Wiley dreamtech India Pvt. Ltd. 2002 ISBN 81-265-0275-4
Publications Publications
Training ProgramsURL’s
© 2009 Wipro Ltd – Internal & Restricted
Thank You
Renu SinghTechnical Lead
renu.singh@wipro.com