Representation Mappers

This module provides a framework for implementing custom representation mapper routines.

Generator-based mappers

These build a mapper from the Pauli images of an algebra’s generators.

Ternary trees

A ternary tree on $N$ nodes defines a fermion-to-qubit encoding of $N$ modes onto $N$ qubits, and the familiar encodings are particular tree shapes: a chain along "Z" is Jordan-Wigner, a chain along "X" is the parity encoding, and the balanced tree attains the optimal Pauli weight $\lceil \log_3(2N+1) \rceil$. Arbitrary trees (including the irregular ones grown from a device’s coupling graph) are built by passing a parent-and-label specification to TernaryTree directly, which is what makes this a framework rather than a fixed set of encodings.

A tree is compiled into a TernaryTreeEncoding once and then passed to any of the ternary-tree mappers in library. Properties derived from the encoding itself, such as TernaryTreeEncoding.total_parity(), are methods on that class.