pymiediff.coreshell.ab_gpu

Contents

pymiediff.coreshell.ab_gpu#

pymiediff.coreshell.ab_gpu(x, y, n, m1, m2, which_jn='stable', precision='double')#

an and bn scattering coefficients - GPU compatible implementation

native torch implementation via bessel upward and downward recurrences

Absorption and scattering of light by small particles. Pg. 183 and Equ. 8.1.

Parameters:

  • x (torch.Tensor) – size parameter (core)

  • y (torch.Tensor) – size parameter (shell)

  • n (torch.Tensor) – orders to compute

  • m1 (torch.Tensor) – relative refractive index of core

  • m2 (torch.Tensor) – relative refractive index of shell

  • which_jn (str) – Which algorithm to use for spherical Bessel of first kind (j_n). must be one of: [‘stable’, ‘fast’], indicating, respectively, continued fractions ratios or standard downward recurrence. Defaults to ‘stable’

  • precision (str, optional) – “single” our “double”. defaults to “double”.

Returns:

result

Return type:

torch.Tensor, torch.Tensor

Contents