Apache Spark provides its users the ability to implement Decision Trees algorithms in a very efficient way, however the output seems to be not so friendly for non-technical users. So this project is an attempt to visualize Collapsible Decision Trees using D3.js, by parsing the nested conditional statements to a JSON format, and using it as dataset for display.
Apache Spark is supported in Zeppelin with Spark interpreter group. In this project also is included Apache Zeppelin notebook template. You can import it and visualize your decision tree horizontally or vertically, which is also collapsible.
- Apache Spark
- Apache Zeppelin
- Python
- HTML
- JavaScript
- D3.js
This is my own implementation and it works great for me, so you are free to use it, modify it, and create magic with it!
In Apache Zeppelin click Import Note and load 'decision-tree-zeppelin-notebook.json' as new notebook.