A few months ago I stumbled upon a fantastic app that showcase the not very well known math functions in Qlik. The app does a great work popularizing math & statistics using default Qlik functions. Ever since I put my hands on the app I wanted to blog about it.
I asked the app's author, Mária Šándorová @JaMajka1 to introduce the app:
I started to develop the app like this the same year I started to work as a full-time Qlik developer. For me it was a good way how to explore mathematical possibilities together with script and expressions syntax of a great tool that was new to me. After that it didn’t have my focus for some time. And then there was a boom of advanced analytics and predictions. Qlik community, different blogs, even my own presentations and Qonnections were full of SSE – possibilities how to utilize mathematical power of R or python for these algorithms.
They are great tools with their advantages in more complex algorithms and a good step forward for Qlik that it can be integrated with them. However, I think that because of all these information and materials about advanced analytics for Qlik in R or Python or anywhere else, we started to underestimate Qlik’s own capabilities.
When someone needs to calculate a correlation, identify outliers, run k-means clustering or test data normality, we really don’t need to use an external tool – we can use Qlik’s default functions as are or script an algorithm within Qlik. I spent days building this app and I still haven’t covered many areas like hypothesis testing or linear regression. Not to mention algorithms you can script within Qlik. So there really are many advanced analytics and statistics possibilities integrated within Qlik itself, too.
I think it’s amazing that we have different options on how and where to calculate something! And I believe that for being able to make a good choice, we need to consider our options and the first step is to know we have it 😊.
So, exploring Qlik’s capabilities in statistics was the first reason to create this app. It defines the content of it. The structure and design of the app is driven by something different. I really like data literacy initiatives and I believe we need to help people become data intelligent and understand their data.
I know, this app is more about mathematics than about context and visualizations, but I think it’s also an integral part of the data intelligent company – even if not necessary important for all users. And what is a better way how to understand the function if not having a simple use case and a generated dataset in an interactive tool?
I love the idea of trying possibilities and seeing the results in the same second. In advanced sections of the app you can select a subset of data in two clicks and thanks to Qlik, everything is recalculated! Select outliers only and see the results of mean and variance … select excluded values and see their results… That’s brilliant for understanding what’s going on!
And if you prefer specific formulas and definitions, they are there, as well 😉.