xnn.cnn.models.spherical.wigner_d#

xnn.cnn.models.spherical.wigner_d(l, beta)[source]#

Wigner d-matrix d^l_{mn}(beta) = <l m| exp(-i beta J_y) |l n>.

Rows and columns run over m, n = -l .. l; the matrix is real and orthogonal, d^l(beta_1) d^l(beta_2) = d^l(beta_1 + beta_2).

Parameters:
  • l (int) – Degree.

  • beta (Tensor) – Angles of any shape.

Returns:

Shape (*beta.shape, 2l + 1, 2l + 1) in float64.

Return type:

Tensor