tree_d_3d.kernels.evaluate_log_quadtree.evaluate_log_quadtree_pyccel

The Pyccel version (CPU) of the evaluate_log_quadtree kernel of the Barnes-Hut and FMM solvers.

evaluate_log_quadtree_cuda.cu next to this file is the CUDA version (GPU) and evaluate_log_quadtree_numpy.py the version for machines without Pyccel, all with the same arguments; the folder’s __init__.py declares the kernel (see :mod:tree_d_3d.kernels). The flat quadtree is described in :mod:tree_d_3d.flat_tree.

Functions

Name	Description
evaluate_log_quadtree	Line-charge potential and field at targets, without k_e.

evaluate_log_quadtreefunction#

View source

def evaluate_log_quadtree(targets: float[:, :], exclude: int64[:], pos: float[:, :], charge: float[:], perm: int64[:], node_center: float[:, :], node_half: float[:], node_start: int64[:], node_end: int64[:], node_child: int64[:], node_qabs: float[:], moment_re: float[:, :], moment_im: float[:, :], theta: float, softening: float, stack: int64[:], phi: float[:], field: float[:, :], status: int64[:]) -> None

Line-charge potential and field at targets, without k_e.

Same traversal and opening rule as :func:evaluate_quadtree. Accepted cells use the order-p expansion, unsoftened: for a target at z, sum q/(z - z_j) = sum_n a_n w**(n+1) and sum q ln(z - z_j) = a_0 ln(z - c) - sum_{n>=1} a_n w**n / n with w = 1 / (z - c). status[0] becomes 1 + a target that met an unsoftened coincident charge (the first, on the CPU).