Why Power BI loves a Star Schema

Are you using a flat table in your Power BI report? You may be missing out! Patrick explores what a STAR SCHEMA is and why you should be using it.


*******************

Want to take your Power BI skills to the next level? We have training courses available to help you with your journey.


*******************
LET'S CONNECT!
*******************


***Gear***

#PowerBI #StarSchema #GuyInACube

[asenkiprov]

Frankly, if it wasn't for your channel, our company probably wouldn't be using Power BI. MS Docs are fine but you guys are sharing that special kind of knowledge that only first-hand experience can foster. Thank you a million times!

[JMcLeodKC]

As a Data Warehouse person, this is the best video that I have ever seen that explains OLAP and Data Warehouse best practices

[rosswaterston]

The simpler DAX and performance are so true. So often when people ask me for help with a measure that's not working well it comes back to data models that don't use the Star Schema.

[thomasivarsson]

I love hearing "Kimball" and "Starschemas -the complete reference", in the video.

[MrDannyBM]

This channel is so straightforward, fast, easy to understand and always has great examples. Thank you so much!

[jackbauir]

Awesome video as always Patrick! Maybe would be usefull for many people if you explain how to break one huge table into a star schema

[LuckyHunter]

I love your presentation style. Your changing tone, expresions, hand gestures, all contribute to making your videos engaging and entertaining. The content is great too, you're my first port of call when I have a Power BI question.

[ariarc]

I agree with Asen Kripov (the previous commenter). If it wasn't for you and your channel, I probably wouldn't be using Power BI. You have a very friendly way of bringing complex, dry material to life. I've only been using it for two days, and am already making very useful visualizations and reports. In a few days I'll have a portfolio set up to pitch the product to my boss and have him buy all the necessary subscriptions. Thanks for your work (which you don't make it look like work, but rather fun time) !

[the_empress_dej]

This is best explanation I've seen on star schema. Thank you!

[dylanbingham]

I might be getting hired onto a helpdesk that specifically handles Power Bi issues, these videos are invaluable. Especially for my interviews. Thank you.

[AyyMz]

Great video! I popped this on while driving home and the concept makes perfect sense (coming from zero experience using star schemas).

[Bharath_PBI]

Thank you. I have been thinking about it on the same lines until recently came to know that star schemas doesn't scale for large models (especially dimensions with 100s of millions of rows) and suggestion received from memory footprints is to flatten the tables. For many other reasons Star schema works best definitely.

[olusolaokuboyejo]

Thanks Patrick. I am very new to Power BI/Data Modelling, and you made my day by making this tutorial easy to understand. Thank you.

[RonaldPostelmans]

Hi Patrick, i did not know that the differences are so high. thanks for that simple explanation.

[matthewburkett]

Thanks for another awesome video Patrick! Would be awesome if you could also show us the best way to turn a flat file into a star schema model.

[reuvmmh]

This is the only video that convinced me to use star schema. Thanks a lot!!!

[drummerboieva]

Wonderful explaination.... my team improved performance of refresh from nearly always fail due to sheer volumetry of data to 99% pass after changing our data model

[koenverbeeck]

Additional benefit from star schema (and is usually my main argument vs a tool like Tableau): if you have one big table some measures might give unexpected results. (Unexpected results are interpreted by the business as incorrect results :)) Suppose you have your gigantic sales table with the customer data alongside it. Now you want to calculate the average age of the customer. whoops, that will be incorrect since some customers will have made multiple sales and thus their age is repeated as well. So you get a weighted average instead. While if you have the customer data separate in a dimension, calculating average age is easy-peazy. This ties in with your argument of easier DAX, because you can fix the average age measure in the one big table, but it's more complex.

[DarkOrint]

This is the single most important thing that changes basic power bi report design into real workable models. Partitioning data is also a must for this. Dimensions can wide lots of cols but dont tend to be as big as fact tables many rows. So not having large facts with text based data is a must, aggregation data only in facts, slice and dice by cols in dimensions.

[christopherthoms]

Thank you for this! I've known that star schema data models are considered best practice in Power BI since I started learning how to use it about 1 year ago. However, I could never fully articulate exactly why. This video helps, giving a great, pithy explanation. Keep up the great work!