FLOUR

flour is a Python library for really fast chemical file reading and writing.

---

Table of contents

Installation

pip install flour

.cube files

import numpy as np
import flour

# Write the cube file.
flour.write_cube(
    path="molecule.cube",
    title1="The first line of the cube file",
    title2="The second line of the cube file",
    # The atomic number of each atom.
    atoms=np.array([1, 35], dtype=np.uint8),
    # The charge on each atom.
    charges=np.array([0., 0.]),
    # The position matrix of the molecule.
    positions=np.array([
        [-0.5, 0., 0.],
        [1., 0., 0.],
    ]),
    # The origin of the voxel grid.
    voxel_origin=np.array([-4.5, -4.5, -4.5]),
    # The a, b, c vectors of a single voxel.
    voxel_size=np.array(
        [
            [0.5, 0., 0.],
            [0., 0.5, 0.],
            [0., 0., 0.5],
        ],
    ),
    # A 20 x 20 x 20 voxel grid.
    voxels=np.random.rand(20, 20, 20),
)

# Read the cube file.
cube_data = flour.read_cube("molecule.cube")
cube_data.atoms  # The atomic number of each atom.
cube_data.charges  # The charge of each atom.
cube_data.positions  # The position matrix of the molecule.
cube_data.grid.origin  # The origin of the voxel grid.
cube_data.grid.voxel_size  # The a, b, c vectors of a single voxel.
cube_data.grid.voxels  # The voxel grid.

Developer guide

If you want to develop FLOUR there's a couple of handy things you should know. FLOUR is built using the excellent maturin library. This means that to build the library you run

maturin develop --extras=dev --profile=release

and that's more or less all there is to it. The only other thing to note is that we do have a justfile, which can be used to run various actions on the repository, such as automated checks and building the Docker testing environment. To use the justfile you should grab just.