Overcoming Python Notebooks’ limitations
As ActiveViam’s engineers build prototypes for clients and our Data Lab works on finding actionable insights on our clients’ market, business, and customers, they regularly need to work on large data sets with a hard deadline in mind.
Python notebooks are our go-to tools for these projects: they’re simple and flexible, but they also have several limitations when it comes to visualization and data volume in particular, so we tried to figure out how to improve notebooks to make analysis at the same time easier, more powerful and more collaborative – for our own use at first.
Bringing Python to ActivePivot
We decided that the best way forward was to try to integrate a full OLAP cube – none other than ActivePivot in fact – into Python notebooks and combine it with ActiveUI for visualization. We wanted to integrate ActivePivot with Python notebooks as seamlessly as possible, in order to combine the capabilities of ActivePivot with the notebooks’ ease-of-use.
The result is a Python library that allows data scientists who use notebooks to:
- Explore data with intuitive and dynamic visualization dashboards
- Overcome the limitations of data volumes
- Benefit from multi-dimensional analysis and “what-if” scenario comparison
- Easily share their results with other users at each step of their analysis or prototyping process
Discover more about our new platform for data science in this article by Antoine Chambille, ActiveViam’s Global Head of Research and Development, on Medium.
If you want to try this new library, contact us with subject : “Python Library beta test”.