contract() should accept a Chain as input

[starsfordummies]

currently I'm getting

contract(sergio_mps)
ERROR: MethodError: no method matching getindex(::Chain, ::Symbol, ::Symbol)

quick workaround to return a tensor is to just

Tenet.contract(mps::Chain) = contract(TensorNetwork(mps))

but if we want to reuse the indices, I came up with something like

 Dense(Quantum(TensorNetwork([contract(TensorNetwork(sergio_mps))]), sergio_mps.super.sites))

oh my

[mofeing]

contract on Chain should be now fixed in the PR but... what is the last thing you're trying to achieve? Converting a MPS into a Dense representation? Why?

[starsfordummies]

oh I just thought if it works for a TN, then it should also work for an MPS - it could even be useful for some benchmark at some point
( I'd probably even put a warning when trying to do it with more than 10 legs or so)

[mofeing]

Closing because this works with the latest PR #172

julia> using Tenet
Precompiling Tenet
  1 dependency successfully precompiled in 3 seconds. 50 already precompiled.
[ Info: Precompiling TenetChainRulesCoreExt [4c48fc49-ee8c-5c6e-a164-94f373290340]

julia> mps = rand(Chain, Open, State; n=5, p=2, χ=20)
MPS (inputs=0, outputs=5)

julia> contract(mps)
2×2×2×2×2 Tensor{Float64, 5, Array{Float64, 5}}:
[:, :, 1, 1, 1] =
 -0.476082  -0.530897
 -0.108129  -0.212065

[:, :, 2, 1, 1] =
 -0.315495  -0.115779
 -0.076073  -0.0214035

[:, :, 1, 2, 1] =
 -0.117174   -0.167136
 -0.0260362   0.0109016

[:, :, 2, 2, 1] =
 -0.154203   -0.127011
 -0.0392941  -0.0132862

[:, :, 1, 1, 2] =
 0.0589333   0.00149326
 0.0189144  -0.0870499

[:, :, 2, 1, 2] =
 0.165494    0.230064
 0.0127105  -0.00592115

[:, :, 1, 2, 2] =
 0.113758   -0.0558552
 0.0175383   0.0481059

[:, :, 2, 2, 2] =
 0.163475   0.246799
 0.0260987  0.180637