Skip to content

Visualizes the Random Forest debug string from the MLLib in Spark using D3.js

License

Notifications You must be signed in to change notification settings

ChuckWoodraska/EurekaTrees

Repository files navigation

EurekaTrees

This program was made to visualize MLLib Random Forests. Running this program with a supplied trees file generates visual version of the trees. You can get this from printing out the debug string of the generated model. In python that is: model.toDebugString()

Install:
pip install eurekatrees

Run:
eurekatrees --trees ./sample_files/trees.txt

Output:
The output path for generated files can be specified with the output_path switch which defaults to the current working dir. eurekatrees --trees ./sample_files/trees.txt --output_path .

If you have a csv with the names of your features you can run that command with the columns switch.
eurekatrees --trees ./sample_files/trees.txt --columns ./sample_files/columns.csv

The output are a bunch of HTML files. It auto opens home.html in your default browser, but you can also open up the file home.html allowing you to easily navigate to each of your trees.

Important Note:
The last change that was made splays out branches further which means more scrolling for right now, but I am looking into d3 canvas zooming to fit it to the page then allow the user to move around inside of the canvas. If anybody has any experience with doing this feel free to contribute or pass along any tips. Thanks!

Example Tree: Example Tree

Credit D3.js tree example from http://bl.ocks.org/d3noob/8326869 was a great help.

About

Visualizes the Random Forest debug string from the MLLib in Spark using D3.js

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Contributors 3

  •  
  •  
  •