periodic non-equidistant grid

[Chao-Chin Yang]

Dear All,

I was asked to send this note to pencil-code-discuss. So here it goes.

I would like to propose a fundamental change to the layout of a periodic non-equidistant grid.

For a periodic non-equidistant grid, current paradigm is to set xi = [0, 1, 2, …, n-1] and x_1 = x(xi=0) = x0 + 0.5 * dx. In this setup, however, the grid would not be symmetric with respect to the origin. For example, if we let x = 0 in the center of the computational domain and set x_star = 0, x_n = x(xi=n-1) is not equal to x1 - 0.5 * dx.

I would presume a symmetric grid is preferable to most people. To achieve this, simply setting xi = [0.5, 1.5, 2.5, …, n-0.5] and x(0) = x0 would do the trick. And this is what I want to modify the module Grid.

Please let me know if you object to this move. If there is no objection, I will change the code and adjust the auto-tests if any results are off, probably sometime next week.

Sincerely,

Chao-Chin Yang
Postdoctoral Scholar
Department of Astronomy and Astrophysics
University of California, Santa Cruz
U.S.A.