Choi: the CJ matrix of a completely positive map is positive semidefinite
ProvedOCB2012.cj_posSemidef_of_completelyPositivelinear-algebraquantum-information
Let be a linear map, with finite. If is completely positive, then its CJ matrix is positive semidefinite:
Here is the Choi–Jamiołkowski (CJ) matrix of the linear map , with unnormalized and the full transpose (OCB p. 3). Entrywise, with indices ordered ,
Complete positivity means that maps positive semidefinite matrices on to positive semidefinite matrices, for every ancilla dimension (the ancilla is the second tensor factor, as in PeresTerno.IsCompletelyPositive). The proof applies this with to the unnormalized maximally entangled projector, and uses that transposition and reindexing preserve positivity. This is one direction of Choi's theorem; the paper uses it in the CPTP characterization on p. 4.
Preamble
import Mathlib import Definitions.Def_PeresTerno_kraus_basics import Definitions.Def_OCB2012_defs import Definitions.Def_OCB2012_cj open Matrix open scoped Kronecker ComplexOrder
Formal statement
namespace OCB2012
theorem cj_posSemidef_of_completelyPositive {x1 x2 : Type*} [Fintype x1] [Fintype x2]
[DecidableEq x1] [DecidableEq x2] (Φ : Matrix x1 x1 ℂ →ₗ[ℂ] Matrix x2 x2 ℂ)
(hΦ : PeresTerno.IsCompletelyPositive Φ) : (cjMatrix Φ).PosSemidef := by sorry
end OCB2012Source
O. Oreshkov, F. Costa, C. Brukner, Quantum correlations with no causal order, Nat. Commun. 3, 1092 (2012), https://arxiv.org/abs/1105.4464v3, p. 4 (CPTP iff M^{A1A2} >= 0 and Tr_{A2} M^{A1A2} = 1), the direction CP => M >= 0; M.-D. Choi, Completely positive linear maps on complex matrices, Linear Algebra Appl. 10, 285-290 (1975), https://doi.org/10.1016/0024-3795(75)90075-0, main theorem (complete positivity <=> positivity of the Choi matrix (Phi(E_ij))_{ij}); stated there for the untransposed Choi matrix, and transposition preserves positivity