Friday, May 1, 2020

4 pillars of Power BI Solutions

Hi All,

On 28 Apr 2020, I spoke on the Power BI SQL PASS community event.

Meetup Event: SG SQL PASS Chapter Meetup - April 2020
Slide Deck and resources (Excel and sample PBIX files) 

There are 4 key pillars of Power BI Solution:





  1. Transform Data

    • Our data will be always raw data that needs to massing for the reports.
    • Data has to into shape and we need to perform a couple of data transformations. 
    • Power BI tries to attempt data types during data transformations. But it may not correct.

  2.  Data Modelling



    • In this phase, we optimizing the tables and columns. 
    • Segmenting common dimensions of your data & build relationships between tables. 
    • Having an effective data model saves so much time
  3. DAX calculations 

    • Formula language that is used in Power BI. 
    • Not only used by PBI but other Microsoft Products. 
    • The primary tool to build Power BI solutions
  4. Report & Visuals

    • This MUST be the last step. You skip the above 3 steps, your BI solutions won't be a success. 
    • More time should be spent on the data model and DAX. 
    • Use appropriate chart / visual for data & STORY you are telling 
    • Key Information STAND OUT
    • Clear & simple title / headings.
    • Understand within 15-20 seconds
  
In this blog post, we will see how to use a sample excel file which can be treated as RAW DATA which as follows data

  1. Sales  - Fact Table
    • It is a transactional table with 7000+ rows.  
    • It has columns such as OrderNumber, OrderDate, Customer Information, Currency, Delivery Information, etc.


       
  2. US Region  - Lookup Table
    • Information about all US regions
  3. Customer List  - Lookup Table

    • Information about all customer and Distribution channel.

  4. Product List  - Lookup Table
    • Information about Products

Customer requirements from this report

  1. KPIs for
    • Total Sales. 
    • The number of units sold 
    • The number unique Order. 
    • Average Sales per day
  2.  Charts about
    • Sales by the Distribution Channel 
    • Sales by the Warehouse code
    • Sales by Region 
    • Sales bu date
  3. Country Map 
    • Sales by regions 
 
We use this Excel file and open Power BI. 


Follow the video where I have how to perform data clean up and data model. 


I will show the DAX and Visual in this blogpost. 

  • Create a simple textbox  Reporting canvas


  • We create a table to store all our DAX formulas. Select Enter Data in the Report 

  • Name this Empty Table as Key Measure


  • In this empty table create  a new measure




  • Name this 1st measure as Sales 
    Sales = SUM( Sales[LineTotal] )






  • We have created the Date table that I have explained in my video.
  •  Drag the Date table.



  • Change from Date Hierarchy 
     

  • Change to Matrix Visual 


  • Drag the Sales in Power BI Desktop



  • Change Visual to Stacked Column Chart. 


  • Drag the Sales Measure and convert it into card visual.


  •  Drag DistributionChannel from consumer list.

  • in our Data model, we have a relationship with the Sales and Consumer table. 
    We can still get DistributionChannel as it is part of the relationship.

  • Drag Sales Measure in this table.

  • Convert to visual to a donut chart.

  • Copy paste this visual twice. It really quickens the report development.

  • 2nd donut chart drag Warehouse code

  • 3rd donut chart drag Region code


  • Drag Map visual and Name field from US Region



    DAX Calculations


  • Create a 2nd measure called Unit Sold. 

    Units Sold = SUM(Sales[OrderQuantity])


    Make it as a Card Visual

  • Copy paste that card visual twice
  • Create 3rd measure called Unique Order

    Unique Order = COUNTROWS(Sales)


    The numbers will change when you any visuals.

  • Create the last and 4th measure called Average Sales Per Day

    Average Sales Per Day = AVERAGEX('Date', [Sales])
  • Adding Year's slicers.



    Add Year slicer. 





I hope you enjoyed this blog-post. 

--aaroh 



 

No comments:

Low Code Reimagined with AI + Copilot Pitch Deck - Copy Copilot day (Virtual) - 2023

 Hi All,  I presneded a session at Pune UG on Low Code Reimagined with AI + Copilot Pitch Deck.  Video is at this address  https://www.youtu...