Getting Started with Power BI

Today we were introduced to Power BI by our Data School coach, Natalia Miteva. this was just Day 1 of a 2-day introduction to Power BI. It was a mixed feeling to finally work on a product that I had heard about in my past jobs. Most of the firms that I had worked with had Office 365 subscriptions and Power BI was the tool used for data analysis.

TLDR

Having been trained in the best in class tools such as Alteryx and Tableau in the Data School, we were quick to judge this too harshly. The way I see Power BI’s capabilities, it lies in data preparation (it has quite cool features especially profiling, transforms and modelling) and quick chart creation. I could see how the way Microsoft tried to add more power and ability to complex transformation using code – DAX and M Query. It felt the same way it was with VBA and Excel. On one hand, it became powerful but increasing complex using VBA code.

Slice and Dice its abilities

Caveat: This is just feedback after using Power BI for a day. I am sure Power BI might continue to surprise me even though I am a big fan of Alteryx.

Pros:

  1. Get Data: There is a long list of data sources you can connect. Especially web-scraping is a breeze. It can easily find HTML tables. It just reduces a lot of work by avoiding using regex expressions in the case of Python/Alteryx to do the same job.
  2. Query Editor: It’s easy to profile data and quickly see your data and make transforms. Tableau Prep has a similar feature and I understand Query Editor had these features quite a while ago. One can easily make changes to data type and view the distribution of the data.

DAX: As per Microsoft documentation, Data Analysis Expressions (DAX) is a library of functions and operators that can be combined to build formulas and expressions in Power BI, Analysis Services, and Power Pivot in Excel data models. It was brought in in 2009 and seems to be a functional language with a lot of useful functions.

Cons:

    1. DAX: DAX finds its place here as I would deem it difficult to understand. It can help achieve a lot of calculations if you are familiar with the syntax. This is to be taken with a pinch of salt. As I read material on DAX, one can write short functions and do complex calculations. However, it is difficult to master and understand. Most of us were left puzzled trying to understand how it works when we started using it. There seems to be a long learning curve for it.
    2. Performance: While we never worried about the size of the files that were cleaned in Alteryx or even visualizations in Tableau, Power BI started slowing down as we built visualizations and added transformation.
    3. Custom Charts: It is not as easy to build custom charts Tableau enables a lot of flexibility and probably frees the inner “artisan” in you. (Artisan is also an idiosyncratically named role that Alteryx Server assigns to developers, however, I think it is more appropriate to more creative Tableau users).
    4. Changes: It does keep track of changes done as part of transformations. Each transformation is recorded. if you click on the change, the GUI shows the actual data and transformation used. However, this becomes a bit clunky to understand if there are a lot of steps. I still the workflow type interface of Alteryx. It is quite easy to understand one’s workflow (after a while) or even others’ workflows.Power BI - Applied Steps

Conclusion

As a premature opinion on a tool that is around for many years, I feel Power BI is a good tool to learn in addition to Alteryx and Tableau. I am grateful that it is a part of the curriculum at the Data School Down Under.