spleaf.cov.Cov.solveLT_back#
- Cov.solveLT_back(grad_x)#
Backward propagation of the gradient for
solveLT()
.Propagate the gradient of a function with respect to x, to its gradient with respect to y and to the components of \(L\) (U, W, phi, G).
Use
cholesky_back()
to propagate the gradient to the initial components of the matrix (A, U, V, phi, F).- Parameters:
- grad_x(n,) ndarray
Gradient of the function with respect to x.
- Returns:
- grad_y(n,) ndarray
Gradient of the function with respect to y.