Skip to content

Latest commit

 

History

History
143 lines (99 loc) · 4.95 KB

README.md

File metadata and controls

143 lines (99 loc) · 4.95 KB

HiPlot - High dimensional Interactive Plotting

Logo

CI Release License: MIT PyPI version PyPI downloads Open In Colab

Community-maintained fork: This is a community-maintained fork of Facebook Research's HiPlot, which has been archived. We aim to keep the project alive with bug fixes, security updates, and new features.

HiPlot is a lightweight interactive visualization tool to help AI researchers discover correlations and patterns in high-dimensional data using parallel plots and other graphical ways to represent information.

Try a demo now with sweep data or upload your CSV or Open In Colab

There are several modes to HiPlot:

  • As a web-server (if your data is a CSV for instance)
  • In a jupyter notebook (to visualize python data), or in Streamlit apps
  • In CLI to render standalone HTML

Quick Start

# Render a CSV to interactive HTML (no install needed)
uvx hiplot-mm data.csv > output.html

# Or start an interactive server
uvx --from 'hiplot-mm[server]' hiplot --port 8765

Installation

# Core package (HTML export, CLI rendering)
pip install hiplot-mm

# With Jupyter notebook support
pip install hiplot-mm[notebook]

# With web server support (hiplot command)
pip install hiplot-mm[server]

# With Streamlit support
pip install hiplot-mm[streamlit]

# Everything
pip install hiplot-mm[all]

If you have a Jupyter notebook, you can get started with something as simple as:

import hiplot as hip
data = [{'dropout':0.1, 'lr': 0.001, 'loss': 10.0, 'optimizer': 'SGD'},
        {'dropout':0.15, 'lr': 0.01, 'loss': 3.5, 'optimizer': 'Adam'},
        {'dropout':0.3, 'lr': 0.1, 'loss': 4.5, 'optimizer': 'Adam'}]
hip.Experiment.from_iterable(data).display()

See the live result

Result

Links

Development

To build from source:

# Install frontend dependencies
bun install

# Install default contributor dependencies (dev/test)
uv sync

# Build JavaScript bundles
bun run build

# Build Python package
uv build

# Or use the all-in-one build script
./build.sh

Common contributor tasks:

# Run Python tests
uv run pytest

# Build docs (docs deps resolved on-demand)
uv run --no-default-groups --group docs sphinx-build -b html docs docs/_build/html

Output directories:

  • npm-dist/ - NPM package artifacts
  • dist/ - Python wheel and sdist
  • hiplot/static/built/ - JS bundle included in Python package

Run the dev server:

uv run --extra server hiplot --port 8765

Citing

@misc{hiplot,
    author = {Haziza, D. and Rapin, J. and Synnaeve, G.},
    title = {{Hiplot, interactive high-dimensionality plots}},
    year = {2020},
    publisher = {GitHub},
    journal = {GitHub repository},
    howpublished = {\url{https://github.com/facebookresearch/hiplot}},
}

Credits

Inspired by and based on code from Kai Chang, Mike Bostock and Jason Davies.

External contributors (please add your name when you submit your first pull request):

License

HiPlot is MIT licensed, as found in the LICENSE file.