Why do I have to initialize the deformation gradient with zeros?

Why do I have to initialize the deformation gradient with zeros, add the ones to the diagonal terms and finally add the displacement gradient? I mean I tried to write down

F = grad(u)
for i in range(3):
    F[i, i] += 1.

instead of

F = zeros_like(grad(u))
for i in range(3):
    F[i, i] += 1.
F += grad(u)

but this seems to give a different result. Honestly, I do not understand why. Do you have any explanation for that? (I took the code snippet from example 36)

In addition to that,I find the different import statements for math helpers a bit confusing - although I understand why they are necessary. I mean these imports:

from numpy import (einsum, linalg as nla, zeros,
                   zeros_like, concatenate, split as npsplit,
                   hstack, abs as npabs, arange, sqrt)
from skfem.helpers import grad, transpose, det, inv

In the beginning I did not understand why I can’t use det and inv from numpy.linalg but then I realized the special structure of the arrays (3,3,n,m). I think this shape is used in order to obtain a reasonable performance with Python/NumPy 😄

Anyway, I’ll experiment with the examples and I hope Github Issues are the right place to ask. Thanks and with best regards, Andreas

[kinnala]

This is a classical mistake when using numpy. I did not notice it first but now it’s obvious.

One should write

F = grad(u).copy()

instead. Otherwise you’ll modify the contents of w.

[kinnala]

In the beginning I did not understand why I can’t use det and inv from numpy.linalg but then I realized the special structure of the arrays (3,3,n,m). I think this shape is used in order to obtain a reasonable performance with Python/NumPy

This is correct. Here is some information related to the forms: https://scikit-fem.readthedocs.io/en/latest/forms.html#helpers-are-useful-but-not-necessary