We establish a 1-1 correspondence between Valiant's character theory of matchgate/matchcircuit [14] and his signature theory of planar-matchgate/matchgrid [16], thus unifying the two theories in expressibility. In [5], we had established a complete characterization of general matchgates, in terms of a set of useful Grassmann-Pl"ucker identities. With this correspondence, we give a corresponding set of identities which completely characterizes planar-matchgates and their signatures. Applying this characterization we prove some negative results for holographic algorithms. On the positive side, we also give a polynomial time algorithm for a simultaneous node-edge deletion problem, using holographic algorithms. Finally we give characterizations of symmetric signatures realizable in the Hadamard basis.
Valiant has proposed a new theory of algorithmic computation based on perfect matchings and Pfaffians. We study the properties of matchgates—the basic building blocks in this new theory. We give a set of algebraic identities which completely characterizes these objects for arbitrary numbers of inputs and outputs. These identities are derived from Grassmann-Plücker identities. The 4 by 4 matchgate character matrices are of particular interest. These were used in Valiant's classical simulation of a fragment of quantum computations. For these 4 by 4 matchgates, we use Jacobi's theorem on compound matrices to prove that the invertible matchgate matrices form a multiplicative group. Our results can also be expressed in the theory of Holographic Algorithms in terms of realizable standard signatures. These results are useful in establishing limitations on the ultimate capabilities of Valiant's theory of matchgate computations and Holographic Algorithms.
We exhibit a strong connection between the matchgate formalism introduced by Valiant and the ZW-calculus of Coecke and Kissinger. This connection provides a natural compositional framework for matchgate theory as well as a direct combinatorial interpretation of the diagrams of ZW-calculus through the perfect matchings of their underlying graphs. We identify a precise fragment of ZW-calculus, the planar W-calculus, that we prove to be complete and universal for matchgates, that are linear maps satisfying the matchgate identities. Computing scalars of the planar W-calculus corresponds to counting perfect matchings of planar graphs, and so can be carried in polynomial time using the FKT algorithm, making the planar W-calculus an efficiently simulable fragment of the ZW-calculus, in a similar way that the Clifford fragment is for ZX-calculus. This work opens new directions for the investigation of the combinatorial properties of ZW-calculus as well as the study of perfect matching counting through compositional diagrammatical technics.
The computational function of a matchgate is represented by its character matrix. In this article, we show that all nonsingular character matrices are closed under matrix inverse operation, so that for every
k, the nonsingular character matrices of
k-bit matchgates form a group, extending the recent work of Cai and Choudhary (2006) of the same result for the case of
k=2, and that the single and the two-bit matchgates are universal for matchcircuits, answering a question of Valiant (2002).
It remains an open question whether the apparent additional power of quantum computation derives inherently from quantum mechanics, or merely from the flexibility obtained by "lifting" Boolean functions to linear operators and evaluating their composition cleverly. Holographic algorithms provide a useful avenue for exploring this question. We describe a new, simplified construction of holographic algorithms in terms of Pfaffian circuits. Novel proofs of some key results are provided, and we extend the approach of [34] to nonsymmetric, odd, and homogenized signatures, circuits, and various models of execution flow. This shows our approach is as powerful as the matchgate approach. Holographic algorithms provide in general
O(nωp) time algorithms, where
ωp is the order of Pfaffian evaluation in the ring of interest (with
1.19≤ωp≤3 depending on the ring) and
n is the number of inclusions of variables into clauses. Our approach often requires just the evaluation of an
n×n Pfaffian, and at most needs an additional two rows per gate, whereas the matchgate approach is quartic in the arity of the largest gate. We give examples (even before any change of basis) including efficient algorithms for certain lattice path problems and an
O(nωp) algorithm for evaluation of Tutte polynomials of lattice path matroids. Finally we comment on some of the geometric considerations in analyzing Pfaffian circuits under arbitrary basis change. Connections are made to the sum-product algorithm, classical simulation of quantum computation, and SLOCC equivalent entangled states.