netket.optimizer.solver.svd

Contents

netket.optimizer.solver.svd#

netket.optimizer.solver.svd(A, b, *, rcond=None, x0=None)[source]#

Solve the linear system using Singular Value Decomposition. The diagonal shift on the matrix should be 0.

Internally uses jax.numpy.linalg.lstsq().

Note

If you pass only keyword arguments, this solver will directly create a partial capturing them.

Parameters:
  • A – the matrix A in Ax=b

  • b – the vector b in Ax=b

  • rcond – The condition number