QuantEl

A library for performing state-specific electronic structure and testing new methods.

Any questions should be directed to Hugh Burton (hgaburton[@]gmail.com)

Installation Instructions

Prerequisites

QuantEl requires a number of external libraries. Most of these can be easily installed using a conda environment. The C++ dependencies (fmt, libint2, and armadillo) are fetched and set up automatically by CMake.

Option A — conda package (recommended)

Build and install quantel as a conda package into your base or target environment:

conda build conda-recipe/
conda install --use-local quantel

This handles all C++ and Python dependencies automatically.

Option B — development install

1. Create the conda environment

conda env create -f environment.yml
conda activate quantel

This installs all required build tools (cmake, pybind11, boost, eigen, llvm-openmp) and Python packages (numpy, scipy, h5py, pandas, pyscf, etc.).

2. PyGNME

The nonorthogonal matrix elements in QuantEl require pygnme. Install it after the conda environment is active (the --no-build-isolation flag is required so the build can find the conda-installed numpy headers):

pip install --no-build-isolation git+https://github.com/hgaburton/pygnme.git

3. Build the C++ extension

All C++ dependencies (fmt, libint2, armadillo) are provided by the conda environment.

mkdir build && cd build
cmake ../
make -j4 install
cd ../

4. Add to your Python path

export PYTHONPATH=~/PATH/TO/QUANTEL:$PYTHONPATH

Linux note — OpenBLAS and OpenMP

On Linux, OpenBLAS (the default conda BLAS) is compiled with pthreads, which conflicts with quantel's OpenMP parallelism and produces a runtime warning. The environment.yml sets OPENBLAS_NUM_THREADS=1 automatically when the environment is created. If you already have the environment, apply it manually:

conda env config vars set OPENBLAS_NUM_THREADS=1 -n quantel
conda activate quantel