WELCOME TO TECH | IMMERSION

Track: SQL/BI

PowerPivot with Excel 2010Presenter: Jeff Jones

Outline

o BI EcoSystem

o PowerPivot for Excel

– What, Why, Who?

o Using PowerPivot

o New Powerful Formulas

o Using DAX Functions

o PowerPivot Deep Dive

– How do they do it?

o Future Directions

PowerPivot in the BI Ecosystem

Business Data

ETL

Cloud Data

Real-Time Event Data

Intranet Portal

Data Warehouse

Enterprise Reporting

Self-Service Spreadsheet Analysis

Data Mining

Multidimensional Analysis

Integrated Reporting

Spreadsheet Collaboration

Scorecards and Dashboards

PowerPivot for Excel

o What?– Tool for Information

Workers

– Leverages existing knowledge in Excel

– Free Add-in into Excel 2010

– Supports analysis of large, related datasets

– Provides a local Analysis Services engine

PowerPivot for Excel

o PowerPivot BreakThroughs:– Large datasets, small

files and fast access

– Create “Analysis Mashup” without VLOOKUP

– Professional-grade formulas (new DAX functions)

– Import data from anywhere

PowerPivot for Excel

o Why?

– Need BI applications

quicker

– Need users to build

their own BI

applications easier

– Need easy

deployment

enterprise-wide when

necessary

PowerPivot for Excel

o Who?– Excel Guru’s

• Business Analyst

• Marketers

• Business users that have become Excel geeks

– Technical Data Guru’s• Database developers

• BI Developers

• Technical people that focus on data analysis applications

Using PowerPivot

o Install PowerPivot add-in

o Open PowerPivot

o Select and Import data

o Define Relationships

o Define PivotTable

o Add calculations

o Create Named Sets

o Save file

Data Connection Options

o Enables information workers to

access data across multiple

heterogeneous systems

o Provides a wide range of

connection options, including

connectivity to third-party

products

o Enables you to use data feeds as

data sources

o Enables portability because data

configuration information is stored

in the workbook

o To access Sharepoint Lists from

PowerPivot install Data Services

Update for .NET 3.5 SP1

Data Relationships

o Automatically recognize relationships based on foreign keys

o Manually create relationships when they are not explicitly defined

Linked Tables

o Link to tables in

workbooks in Excel

o Update the table in

PowerPivot as the

linked table in

Excel changes

o Update

automatically or

manually

PivotTable Tables and PivotChart

Charts in PowerPivot for Excel

o Add PivotChart charts

and PivotTable tables

with drill-down

capabilities

o Use the task pane to

add values, labels, and

filters

o Use extensive

formatting options to

create visually

impressive

applications

PowerPivot Slicers

o Provide a quick way

to filter the data that

is displayed in

charts and tables

o Add vertical or

horizontal slicers to

suit the layout of the

application

Using DAX Functions

o Create Calculated Columns in PowerPivot tables to cleanse and extend

o Create new Measures in Excel PivotTables for advanced analysis

o Works with relational data

o Performs dynamic aggregations

o Looks similar to Excel formulas

– Except they reference rows/columns not ranges

o Does not replace SSAS MDX

DAX Function Categories

o Date and Time (EOMonth, WeekNum)

o Filter and Value (Calculate, All, Related))

o Logical Test (True, False)

o Math (Round, Sqrt)

o Statistical (AverageX, CountRows)

o Text (Lower, Mid, Search)

o Time Intelligence (StartOfMonth, ParallelPeriod)

Aggregation Functions

o Functions to calculate averages

o Functions to count values

o Functions to return maximum and minimum values

o Functions to return sum totals

=AVERAGEX(InternetSales, InternetSales[Freight]+

InternetSales[TaxAmt])

=COUNTAX(FILTER('Reseller',[Status]="Active"),[Phone])

=MINX(FILTER(InternetSales, [SalesTerritoryKey] =

5),[Freight])

SUMX(FILTER(InternetSales,

InternetSales[SalesTerritoryID]=5),[Freight])

Filters, Values, and Relationships

Functions

o RELATED and RELATEDTABLE

o FILTER

o ALL and ALLEXCEPT

o CALCULATE and CALCULATETABLE

o VALUES and DISTINCT

=SUMX(ALL(ResellerSales_USD), ResellerSales_USD[SalesAmount_USD])

=CALCULATE(SUM(ResellerSales[SalesAmount]),Reseller[ProductLine] =

"Mountain")

=COUNTROWS(VALUES(Reseller[ResellerName]))

=RELATED(Products[EnglishProductName])

=SUMX(FILTER(Resellers,[ProductLine] = "Mountain"),

ResellerSales[ExtendedAmount])

Calculated Columns and Measures

o Calculated columns – Define and populate a new column based on calculations using values in

other columns

– Calculated columns available to PivotTable tables and PivotChart charts

–

o Measures– Added to Pivot Tables and Pivot Charts

– PowerPivot creates implicit measures automatically

– DAX allows you to create explicit measures to perform more complex

analysis

=RELATED(dimSalesTerritory[SalesTerritoryCountry]) & ",

" & RELATED(dimSalesTerritory[SalesTerritoryRegion])

=CALCULATE(SUM(FactInternetSalesPowerPivot[SalesAmount]),ALL

(dimProductSubcategory[EnglishProductSubcategoryName]))

PowerPivot Deep Dive

o Data exposed as tabular, underlying model is UDM (Analysis Services)

o VertiPaq engine stores data as columns, not rows (Column Store)

o Significant compression and fast retrieval

PowerPivot Deep Dive

o VertiPaq runs

in-process with Excel

o Uses AMO and

ADOMD.Net API’s

o Excel queries data

using MDX

o Uses $EMBEDDED$

data source

connection

Future Directions

o Project “Crescent”– Uses the new BI Semantic Model

– KPIs

– Perspectives

– Multi-column relationships

– Hierarchies

– Parent-child hierarchies

– Many-to-many relationships

– Drillthrough

– Role-based security model using Active Directory

– Row and column level security to secure data at the deepest level

Summary

o PowerPivot for Excel and how it fits in the

BI EcoSystem

o Loading, managing and accessing data in

PowerPivot

o Using DAX functions for sophisticated

analysis

o PowerPivot deep dive on technology

o PowerPivot future direction

Thank you.