PyHermes#

PyHermes is the Python implementation of Hermes, an in situ multiresolution framework for cosmic statistics. A particle catalogue is projected once into a reusable scaling-function-coefficient field (SFCField). Smoothing, binning, multipole projection, and differential operations are then expressed as windows acting on that field.

This field–window language supports count-in-cell statistics, isotropic and anisotropic 2PCFs, standard and multipole 3PCFs, marked and weighted statistics, and derived physical fields such as velocity divergence, Newtonian potential, and acceleration.

Project links: GitHub repository | Issue tracker | PyPI package | Hermes paper | Tutorial notebooks

Catalogue, window, MRA field, and task layers in PyHermes

The four-layer PyHermes workflow used in the Hermes paper. Catalogue data and window operators meet in the MRA layer; task objects assemble the resulting filtered fields into statistics or physical diagnostics.#

The shortest route through the documentation is:

  1. Hermes in one page for the algorithm and its scope.

  2. Installation and Quick start for a first run.

  3. Particle input and conversion when adapting your own catalogue.

  4. Build an SFCField and Field and window operations for the two reusable core objects.

  5. Choose an application path: physical fields, Counting, 2PCF, standard 3PCF, or 3PCF multipoles.

The public examples directory is the executable companion to this guide. Its notebooks are tutorials; its scripts and YAML files are production-shaped starting points.

The notebook gallery renders those same tracked notebooks directly in this site, including their saved figures and outputs.

Start here

Field--window concepts

Physical fields and statistics

Notebook tutorials

Reference and validation