installation.md

Installation

QES requires Python 3.10 or later. It supports both CPU and GPU execution (via JAX).

Quick Install (Pip)

If you have the source tree, install from the Python/ subdirectory:

cd Python
python -m pip install .

This gives you the lightweight minimal package. import QES works without pulling in the JAX stack.

Recommended standard install for NQS/VMC:

python -m pip install ".[standard]"

The .[jax] extra remains available as a compatible alias for the same accelerator-oriented stack.

To include all optional dependencies (docs, JAX, tests):

python -m pip install ".[all]"

Developer Install (Editable)

For researchers and developers who modify the code:

1. Clone the repository:

   git clone https://github.com/makskliczkowski/pyqusolver.git
   cd pyqusolver

2. Install in editable mode:

   cd Python
   python -m pip install -e ".[standard,dev]"

   Note: The Python/ subdirectory contains the installable package.

Dependencies

Core dependencies are defined in Python/pyproject.toml and include:

• numpy, scipy, matplotlib
• pandas, sympy, tqdm, numba

Recommended standard extras:

• standard or jax: jax, jaxlib, flax, optax (for NQS/VMC and accelerator-ready workflows)

Additional optional extras:

• hdf5: h5py (for data storage)
• ml: scikit-learn, scikit-image
• all: full stack including JAX, docs, ML, HDF5, and dev tooling

Verification

After installation, verify it works by importing QES:

import QES
print(QES.__version__)

If you need the neural-network workflows, verify the JAX extras separately:

import QES
from QES.NQS import NQS
print("QES and NQS import OK")