GitHub

EDKit.jl

EDKit.jl is a Julia package for exact diagonalization, symmetry-resolved many-body calculations, and selected tensor-network and open-system workflows.

This documentation site is the human-facing manual for the package. It focuses on concepts, workflows, and examples. If you are reading the code directly, the source-level docstrings in src/ are intentionally more exhaustive and more implementation-oriented.

It is built around a simple pattern:

  1. describe local terms once,
  2. choose the basis or symmetry sector you care about,
  3. construct an operator,
  4. use that same object as a linear map, a dense matrix, or a sparse matrix.

What EDKit Covers

EDKit has a few major subsystems:

  • basis construction for full Hilbert spaces, projected subspaces, and symmetry sectors,
  • many-body operator construction from local terms,
  • inter-basis maps and symmetrizers,
  • entanglement and Schmidt-decomposition helpers,
  • ITensor/MPS utilities, including Pauli-space workflows,
  • Lindblad and quadratic-fermion open-system tools.

If you are new to the package, start with Getting Started, then move to the manual pages for the subsystem you need.

Installation

Install the registered package with Julia's package manager:

pkg> add EDKit

Or use the GitHub repository directly:

pkg> add https://github.com/jayren3996/EDKit.jl

Current package compat:

  • Julia 1.9+
  • ITensors.jl 0.7 to 0.9
  • ITensorMPS.jl 0.3

Choose A Path

I want to build and diagonalize many-body Hamiltonians

Start with:

I want to reduce into symmetry sectors

Start with:

I want to work with ITensors, MPS, or Pauli-space representations

Start with:

I want open-system or quadratic-fermion evolution

Start with:

Design Philosophy

EDKit keeps model construction separate from representation choice.

That separation matters in practice:

  • you can stay matrix-free when systems are too large for explicit diagonalization,
  • convert to dense arrays when exact eigensolvers are appropriate,
  • convert to sparse matrices for sparse linear algebra or interoperability,
  • and reuse the same local terms across different bases and symmetry sectors.

Manual Layout

The docs are organized in three layers:

  • the manual explains concepts and workflows for human readers,
  • the worked examples show compact end-to-end use cases,
  • the API reference lists the core functions and types grouped by subsystem,
  • the source docstrings in src/ carry the most detailed function-by-function semantic documentation.

The package is a single Julia module, EDKit, but it contains several functional layers. The Architecture page explains how those layers fit together.