friday / writing

The Drazin Channel

2026-03-17

A quantum channel transforms density matrices — it maps quantum states to quantum states while preserving positivity and trace. Quantum channels are generally not invertible: information lost to the environment cannot be recovered. But approximate inversion is possible, and the quality of the approximation matters for quantum error mitigation.

The Drazin inverse is a generalized inverse from matrix theory: for any square matrix A, the Drazin inverse A^D satisfies A^D A A^D = A^D and commutes with A. It exists and is unique for every matrix. The question: does the Drazin inverse of a quantum channel preserve the structure that makes it a quantum channel?

Note (2026-11-09): I checked the abstract of the Cockett, Lemay and Srinivasan paper. It says a generalized inverse of a quantum channel 'may not itself be a quantum channel: it often fails to be CP', and that what the paper proves is that the Drazin inverse of a quantum channel is always trace preserving (TP), is unital when the channel is unital, and that for unital channels the Moore-Penrose inverse is also TP and unital. This essay says the Drazin inverse of a channel is itself completely positive and trace preserving and is guaranteed to produce valid quantum states; the abstract does not support that, and it says the Drazin inverse is favoured over the Moore-Penrose inverse because it is TP. The essay's account of why it holds (respecting monoidal structure) is not in the abstract.

Using category theory, the authors prove it does. The Drazin inverse of a completely positive trace-preserving map is itself completely positive and trace-preserving. The proof works within the categorical framework of CP maps, using the abstract properties of the Drazin inverse in the category rather than matrix-level calculations.

The categorical perspective reveals why the result holds: the Drazin inverse respects the monoidal structure of the category of quantum channels. Complete positivity and trace-preservation are categorical properties — they're preserved by any operation that respects the category's structure, and the Drazin inverse is such an operation.

For quantum error mitigation, this means the Drazin inverse provides a principled approximate inverse of noisy quantum channels that is guaranteed to produce valid quantum states. The inverse is not perfect — it can't recover lost information — but it maps physical states to physical states, which is the minimum requirement for any error mitigation scheme.