Interactive Demos

Each of the following demos are provided as Jupyter notebooks that are available on GitHub. Binder provides an awesome platform for trying out these notebooks without installing anything whatsoever.

For more computationally intensive tasks, such as training EFNs/PFNs or other neural networks, check out the Examples.

EnergyFlow Demo

The EnergyFlow Demo provides an introduction to using EnergyFlow to compute Energy Flow Polynomials.

EMD Demo

The EMD Demo provides an introduction to using EnergyFlow for computing the Energy Mover's Distance (EMD) between jets.

MOD Jet Demo

The MOD Jet Demo provides an introduction to using CMS Open Data in the MOD HDF5 format via EnergyFlow. Jets from the CMS 2011A Jet Primary Dataset have been processed into this easy-to-use format and are available on Zenodo along with corresponding simulated datasets.

EFM Demo

The EFM Demo provides an introduction to Energy Flow Moments to compute EFPs and verify linear relations among them.

Counting Leafless Multigaphs with Nauty

Though not demonstrating features of the EnergyFlow library, this script uses the graph program Nauty to count leafless multigraphs, a task which relates to EFMs as described in 1911.04491. To run this script, first download Nauty and compile it. Then point the Python script to the Nauty installation directory. Note that running with -d 16 has been known to take half a month.