pytc.legacy.df.isdf_decompose_cholesky

pytc.legacy.df.isdf_decompose_cholesky(rho, n_rank)[source]

ISDF decomposition using pivoted Cholesky, generalized for tensors. :param rho: Input tensor (N_b, N_grid, *trailing_dims) or (N_b, N_grid) :param n_rank: Number of interpolation points

Returns:

Selected columns from rho (N_b, n_rank, *trailing_dims) or (N_b, n_rank) xi: Interpolation coefficients (n_rank, N_grid, *trailing_dims) or (n_rank, N_grid)

Return type:

C

Parameters:
  • rho (numpy.ndarray)

  • n_rank (int)