Learning Python/R VS using SSBI tools (Tableau and Power BI)

Recently, have started developing my Data Analysis & Visualization skills. One of the critical questions asked

why i have to learn R or Paython while i've access to very powerful tools such as Power BI?

I'll be happy to share with you the conclusion have reached from my research.

Let's start stating the key features of each side

Power BI can connect to (complete list at the end):

•    Databases: Relational (SQL) and Non-relational (Mongo DB using driver)
•    Localfiles: ( .xls .csv .sql .json .xml , plain text, , SharePoint Folder)
•    Hosted files: etc
•    Can manage credentials/authentication (For example, windows credentials or End User authentication OAuth)
•    Connect to and extract data (Web content)
•    Use online services to gather data (third-parties like Marketo or GitHub)
•    Seamless connection with Azure (or that’s what Microsoft wants you to believe in)

Think of it this way… PowerBI, Tableau etc are like being a TV weather reporter… You can present the information about the weather very well and you can visualize it in about a way that people understand. You can present the metrics that will help people decide what to do that day - temperature, wind speed, humidity, etc. That's valuable. People need that in the way that businesses need BI: give them the information that they need to make the decisions they know they need to make.

Python and R are a little further up the chain than that. More like being a meteorologist. You study data in a more scientific, way. You find correlations between weather trends and impacts. You build models to predict what’s going to happen based on historical trends. These are programming languages that can be used to build extremely powerful tools to solve more niche and custom needs.

They can connect pretty much with anything. I just want you to notice that if you are using Power BI you can connect and parse a website in seconds (assuming that is written in neat tables and all of that) and in Python/R you would have to write code ( not from zero with Beautiful Soup for Python and ScrapeR for R).

VISUALIZATION

Power BI

PowerBI can visualize data using different visuals and high customizable options for each graph (you can even code your own from scratch There are main concepts here:

• Managers/Analysts/clients nowadays know how to interact with your mobile or web view report with 5 minutes of training or even without it (they can know PBI beforehand). The behavior will remain constant through every report (e. g. cross-highlight data). You just don’t need to teach them how to use your Python or R interface.
• Additional visuals: Visuals Gallery | Microsoft Power BI
• Personalize visuals: Microsoft/PowerBI-visual
• Create your own R to plot graphs! R Visuals in Power BI
• Create your image and just show on your report using a public URL (this can save your life sometimes)

Python and R

You can achieve good results using Python or R. The major downside that they are going to be statical (I’m not talking about useless 3D graphs). If you are presenting a report, some questions will arise sometimes in the middle of a meeting. Like for example, “in that X age group, how many women do we have?” This is solved by one click using Power BI by using dynamic filters and cross-highlighting.

Yes, It's also possible using Python, R and/or javascript but it's not of the shelf and would take time and planning.

 My conclusion:

if you want to focus solely on presenting data and making it interactive, BI is the way to go. If you want to do some really powerful buy probably more niche things with data, a little R/Python coding goes a long long way.

Sources: Quora, Microsoft Power BI community