Do we need increase_scaling_factor and init_scaling?

[marcelmbn]

Really necessary?

For the sake of simplicity, I though about getting rid of the user-accessible variables init_scaling and increase_scaling_factor (see below).

# > Initial coordinate scaling factor. Options: <float>
init_scaling = 3.0
# > Increase in the coordinate scaling factor per trial after check_distance was not met. Options: <float>
increase_scaling_factor = 1.1

If they are never accessed by users (@gasevic, @jonathan-schoeps?) we should maybe remove them from the toml file and from the CLI as they blow up the complexity of the program from the user's perspective. What do you think?

The relevant variables are here called scaling and inc_scaling_factor.

def generate_coordinates(
    at: np.ndarray,
    scaling: float,
    inc_scaling_factor: float = 1.3,
    verbosity: int = 1,
    scale_minimal_distance: float = 0.8,
) -> tuple[np.ndarray, np.ndarray]:
    """
    Generate random coordinates for a molecule.
    """

    # eff_scaling is a deep copy of scaling
    eff_scaling = copy.deepcopy(scaling)
    xyz, ati = generate_random_coordinates(at)
    xyz = xyz * eff_scaling
    # do while check_distances is False
    while not check_distances(xyz, ati, scale_minimal_distance=scale_minimal_distance):
        if verbosity > 1:
            print(
                f"Distance check failed. Increasing expansion factor by {inc_scaling_factor}..."
            )
        xyz, ati = generate_random_coordinates(at)
        eff_scaling = eff_scaling * inc_scaling_factor
        xyz = xyz * eff_scaling

    return xyz, ati


def generate_random_coordinates(at: np.ndarray) -> tuple[np.ndarray, np.ndarray]:
    """
    Generate random coordinates for a molecule.
    """
    atilist: list[int] = []
    xyz = np.zeros((sum(at), 3))
    numatoms = 0
    rng = np.random.default_rng()
    for elem, count in enumerate(at):
        for m in range(count):
            # different rules for hydrogen
            if elem == 0:
                xyz[numatoms + m, :] = rng.random(3) * 3 - 1.5
            else:
                xyz[numatoms + m, :] = rng.random(3) * 2 - 1

            atilist.append(elem)

        numatoms += count

    ati = np.array(atilist, dtype=int)

    return xyz, ati

[marcelmbn]

Btw, I wanted to try the Discussions feature of GH. 😄🎉

[jonathan-schoeps]

I did not touch this values once. I would be fine with it to remove this from the .toml file.

[gasevic]

I also think it is okay to remove this.