Stefano Carrazza

icon indicating a website link https://stefanocarrazza.com

icon company University of Milan icon location Milan icon location Research scientist in computational physics, particle physics, quantum physics and machine learning

Results 104 comments of Stefano Carrazza

Time evolution of mixed states (density matrices)

@DiegoGM91 thanks for reporting this feature, if you have some implementation please open a PR or share the code with us.

Including derivative in parallel L-BFGS-B method

@AdrianPerezSalinas thanks for this issue. I think this is feasible, by computing the analytic gradients manually or automatically (via tensorflow, or other backends).

Including derivative in parallel L-BFGS-B method

Usually, automatic analytic gradients are more efficient than finite differences, which are already computed by the scipy minimize.

Including derivative in parallel L-BFGS-B method

Thanks for the clarification, sorry for the misunderstanding. Yes, this is something which may help, and for sure interesting to have build-in.

Including derivative in parallel L-BFGS-B method

Ok, thanks for this tests, lets see.

Including derivative in parallel L-BFGS-B method

OK and how this compares to the numerical derivative for similar configurations?

Trainable parameters for gates with multiple inputs.

@igres26 thanks for reporting. I believe we can have a `trainable` flag which accepts a boolean or a mask.

Expectation value of a Hamiltonian

Do you have references about this feature?

Add generalized depolarizing channel

Thanks @renatomello, we are looking into this.

Quality of life improvements

Thanks @igres26 , that's a quite good list wish list.