c-bowman / tokamesh Goto Github PK

This line in construction.build_central_mesh pads the input boundary by a fixed amount:

This looks like maybe it's supposed to pad by the height of two triangles, i.e. 2 * sqrt(3) * resolution * 0.5? I don't think this affects the validity of the function -- it looks like it would only shift the start of the resulting grid?

On a related note, the docstring for padding_factor in this function says:

    :param padding_factor: \
        A multiplicative factor which defines the minimum distance to the boundary
        such that ``min_distance = padding_factor*scale``. No vertices in the returned
        mesh will be closer to the boundary than ``min_distance``.

Should scale be resolution?

Is it worth investigating using https://docs.scipy.org/doc/scipy/reference/generated/scipy.spatial.Delaunay.html instead of triangle.c?

Doing so would solve the issue of compiling triangle on the fly.