Follow these steps to get the full system up and running.
The most recent master branch commits may break stuff, check out a tagged version (same version for front- and backend) to make sure everything works.
- Setup the backend: Follow the instructions in the backend readme (Setup)
- Setup the frontend: Follow the instructions in the frontend readme or simply
- Install NodeJS if not yet installed
cd frontend
(from the project root)npm i
npm start
Run a batch job
cd backend
(from the project root)source /venv/bin/activate
(if not already activated)python3 batch.py jobs/iris.json
(or use any other batch job file instead ofiris.json
)- Wait until the job has finished (should be less than 2 minutes for
iris.json
)
- Start the backend (if it is not yet running)
cd backend
(if not already here)source /venv/bin/activate
(if not already activated)python3 main.py
- Start the frontend
cd frontend
(from the project root)npm start
- Open a browser and go to localhost:3000
- Load data
- Select a dataset
- Select some classifiers or just keep all selected
- Press the start button
- Wait until the data has loaded
- When the data has loaded, the ranking visualization will be displayed
- Use the frontend
- You can change views at the top
- The Help view contains useful tips and an explanation of all views
- The toolbar (top) provides options depending on the current view, like sorting or coloring
- The status bar (bottom) shows information on data and current filter and highlight
- Some views support zooming and panning via scrolling and dragging
- Most views support highlighting via hovering
Run the most important example batch jobs in one go (should not take more than 20 minutes):
python batch.py jobs/breast_cancer.json jobs/digits.json jobs/iris.json jobs/wine.json jobs/tests/scaling/iris1000.json jobs/tests/multi_label.json
See getting started for a step-by-step guide on plugins and batch jobs.
If you encounter bugs or have ideas, you can mail me, create a text file here or open an issue.