Issue in mps.jl file, line 899

[dhqian]

I encountered an issue in the mps.jl file on line 899, where the code reads:

Li21 *= prime(dag(si[pL21]), !sᵢ)

It appears that the variable si is used, but based on the context, I believe it should actually be psi. This leads to problems when calling the correlation_matrix function with different operators and assigning specific sites.

[mtfishman]

Thanks for pointing that out, do you have an example you can share where that bug is hit so we can fix it and add a regression test?

[dhqian]

An example of the case where the bug hit:

sites = siteinds("Qubit", 10)
psi = random_mps(sites; linkdims=3)
correlation_matrix(psi, "X", "Z", sites=2:2:6)

[emstoudenmire]

Thanks @dhqian. I just made a PR that has a test which reproduces the error, and a patch that fixes it.