Eric Giguère

I updated the krylov page after the merger. The code was not touched since the review.

> The current `.evolve_dm` and `.sys_dims` concepts both seem a bit awkward, so I'd like to suggest the following: > > **For `.evolve_dm`**: > > Replace `.evolve_dm` with a `.unitary`...

The stochastic solver are not thought for a lot of noise operators. In v4, operations are all `CSR @ Dense_1D_array`. In v5, you have more control on which storage is...

We don't use the same code for the sparse addition as in `v4`. v5 sparse addition is quite faster. But I expect it to be slower for 100 additions of...

It will convert the `CSR` to `Dense` then use `add_dense`. `iadd` for sparse is not easy. When none-zero entry are not lining up, the array need to be expanded. Also...

It will be a lot of work to add a COO format, there are a few dozens of functions to implement (add, expect, mul, matmul, ptrace, eigen, kron, ...). We...

Thank you for the suggestions. (2) It should be easy. We could reuse the `parallel_map` as you said. I would make a good first contribution. (1) We need to be...

It's fine to not split the PR in too many chunks. Yes it's easier to review, but I don't want to have it half implemented/merged at release, which is coming...

Either `nnz` return `size_t` or raise an error when allocating matrix with too many elements.

For jax, the issue is with `isherm`, not the casting. We cannot call it in jitted functions since they cannot branch depending on the input data. The casting itself cause...