pith. sign in

arxiv: 2605.19041 · v1 · pith:UO54ZGTXnew · submitted 2026-05-18 · 🧮 math.NA · cs.NA

Recovering Complex Unitary Eigenspaces from Real-Valued Embeddings

Stefanie G\"unther , N. Anders Petersson This is my paper

Pith reviewed 2026-05-20 07:31 UTC · model grok-4.3

classification 🧮 math.NA cs.NA
keywords unitary matriceseigendecompositionreal embeddingsstructured projectionrank-revealing orthonormalizationdegenerate eigenvaluescomplex conjugate pairs
0
0 comments X

The pith

Structured projection followed by rank-revealing orthonormalization recovers the complex unitary eigendecomposition from any real-valued embedding.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper addresses the recovery of a unitary eigendecomposition for a complex unitary matrix when only its real-valued embedding is available from real-arithmetic solvers. Such embeddings split the matrix into real and imaginary blocks, but degenerate eigenvalues or conjugate pairs mix the eigenspaces with contributions from both the matrix and its conjugate. The authors prove that a structured projection applied to these eigenspaces, followed by rank-revealing orthonormalization, always resolves the mixing. This yields the original eigenvalues and a unitary eigenbasis while preserving correct multiplicities. A reader would care because it lets workflows keep using efficient real solvers without losing accuracy on complex unitary problems.

Core claim

We prove that this ambiguity can always be resolved by applying a structured projection to the eigenspaces of the real-valued embedding, followed by a rank-revealing orthonormalization. The resulting procedure recovers the eigenvalues and a unitary eigenbasis for the original unitary matrix, with correct multiplicities of degenerate eigenvalues.

What carries the argument

The structured projection onto eigenspaces of the real-valued embedding, which separates real and imaginary contributions before rank-revealing orthonormalization isolates the unitary eigenbasis.

If this is right

  • The recovery procedure works for arbitrary unitary matrices, including those with degenerate eigenvalues or conjugate pairs.
  • Correct multiplicities are preserved without extra spectral assumptions.
  • The method integrates directly with existing real-arithmetic eigensolvers in scientific computing.
  • Reconstruction remains exact as long as the embedding follows the standard real-imaginary block form.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The projection technique might extend to other structured matrices that admit real embeddings with conjugate symmetry.
  • Numerical experiments on large-scale problems could check how round-off affects the rank-revealing step.
  • Similar separation ideas could apply to recovering eigenstructures in related fields such as quantum information or control theory.

Load-bearing premise

The real-valued embedding uses the standard block construction from the real and imaginary parts of the complex unitary matrix, with eigenspaces that contain separable information.

What would settle it

Take any complex unitary matrix with a degenerate eigenvalue, form its standard real embedding, extract an eigenspace, apply the structured projection and rank-revealing orthonormalization, then verify whether the output basis is unitary and the eigenvalues match the original with correct multiplicity; any mismatch would disprove the recovery guarantee.

read the original abstract

We consider the problem of recovering a unitary eigendecomposition of a complex unitary matrix from that of its embedded real-valued formulation. Such formulations arise naturally in scientific computing workflows that employ real-arithmetic solvers by representing complex matrices in term of their real and imaginary parts. While the reconstruction is trivial when the spectrum of the real-valued embedding is simple, degenerate and/or complex conjugated eigenvalues introduce ambiguities because each eigenspace may include contributions from both the unitary matrix and its complex conjugate. We prove that this ambiguity can always be resolved by applying a structured projection to the eigenspaces of the real-valued embedding, followed by a rank-revealing orthonormalization. The resulting procedure recovers the eigenvalues and an unitary eigenbasis for the original unitary matrix, with correct multiplicities of degenerate eigenvalues.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

0 major / 2 minor

Summary. The manuscript proves that the unitary eigendecomposition of a complex unitary matrix U can be recovered from the eigendecomposition of its standard real block embedding R by applying a structured projection to the real eigenspaces of R followed by rank-revealing orthonormalization. This procedure resolves ambiguities arising from conjugate eigenvalue pairs and degeneracies, recovering the eigenvalues of U together with a unitary eigenbasis and the correct algebraic multiplicities.

Significance. If the central proof holds, the result supplies a parameter-free, constructive algorithm that enables reliable extraction of complex unitary information from real-arithmetic eigensolvers. This is directly useful in scientific computing workflows that avoid complex arithmetic. The handling of general spectra (including degeneracies) without extra assumptions on the spectrum is a clear strength, and the approach is self-contained.

minor comments (2)
  1. A short numerical illustration early in the paper (e.g., a 2-by-2 or 4-by-4 example with a conjugate pair) would help readers see the ambiguity before the general argument.
  2. Clarify in the notation section whether the rank-revealing step returns an orthonormal basis that is unique up to global phase or up to unitary mixing within degenerate subspaces.

Simulated Author's Rebuttal

0 responses · 0 unresolved

We thank the referee for the positive and accurate summary of our manuscript, the recognition of its significance for real-arithmetic workflows, and the recommendation of minor revision. We appreciate the constructive tone of the report.

Circularity Check

0 steps flagged

No significant circularity; self-contained proof

full rationale

The paper's central claim is a mathematical proof that a structured projection onto eigenspaces of the standard real block embedding, followed by rank-revealing orthonormalization, resolves ambiguities from conjugate pairs and degeneracies to recover the exact unitary eigenbasis and eigenvalues of the original complex unitary matrix. This derivation relies on the algebraic properties of the embedding R = [[Re U, -Im U], [Im U, Re U]] as the real representation of a complex-linear map and the fact that unitary matrices are normal (hence diagonalizable over C), without any reduction to fitted parameters, self-definitional loops, or load-bearing self-citations. The procedure is presented as a constructive algorithm derived directly from these properties, making the result self-contained and independent of prior results by the same authors.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The work relies on standard linear algebra properties of unitary matrices and their real embeddings without introducing new free parameters or invented entities.

axioms (1)
  • standard math Unitary matrices satisfy U^* U = I and their real embeddings preserve the spectrum structure up to conjugation.
    Invoked implicitly when discussing the embedding and eigenspace ambiguities.

pith-pipeline@v0.9.0 · 5661 in / 1170 out tokens · 34826 ms · 2026-05-20T07:31:03.236134+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

18 extracted references · 18 canonical work pages

  1. [1]

    , title =

    Freund, Roland W. , title =. SIAM Journal on Scientific and Statistical Computing , volume =. 1992 , doi =

    work page 1992

  2. [2]

    , title =

    Day, David and Heroux, Michael A. , title =. SIAM Journal on Scientific Computing , volume =. 2001 , doi =

    work page 2001

  3. [3]

    Numerical Algorithms , volume=

    A comparison of iterative methods to solve complex valued linear algebraic systems , author=. Numerical Algorithms , volume=. 2014 , publisher=

    work page 2014

  4. [4]

    and Bai, Z

    Anderson, E. and Bai, Z. and Bischof, C. and Blackford, S. and Demmel, J. and Dongarra, J. and Du Croz, J. and Greenbaum, A. and Hammarling, S. and McKenney, A. and Sorensen, D. , title =. 1999 , address =

    work page 1999

  5. [5]

    SIAM review , volume=

    Julia: A fresh approach to numerical computing , author=. SIAM review , volume=. 2017 , publisher=

    work page 2017

  6. [6]

    2024 , url =

    MATLAB version: 24.2 (R2024b) , publisher =. 2024 , url =

    work page 2024

  7. [7]

    Harris, K

    Charles R. Harris and K. Jarrod Millman and St. Array programming with. 2020 , month = sep, journal =. doi:10.1038/s41586-020-2649-2 , url =

    work page doi:10.1038/s41586-020-2649-2 2020

  8. [8]

    http://www.scipy.org/

    Eric Jones and Travis Oliphant and Pearu Peterson and others , title =. 2001-- , url = "http://www.scipy.org/", note =

    work page 2001

  9. [9]

    SIAM Journal on Matrix Analysis and Applications , year =

    Braman, Karen and Byers, Ralph and Mathias, Roy , title =. SIAM Journal on Matrix Analysis and Applications , year =. doi:10.1137/S0895479801384573 , url =

    work page doi:10.1137/s0895479801384573

  10. [10]

    Francis, J. G. F. , title =. The Computer Journal , volume =. 1961 , month =. doi:10.1093/comjnl/4.3.265 , url =

    work page doi:10.1093/comjnl/4.3.265 1961

  11. [11]

    2013 , publisher=

    Matrix Computations , author=. 2013 , publisher=

    work page 2013

  12. [12]

    and Chuang, Isaac L

    Nielsen, Michael A. and Chuang, Isaac L. , biburl =

    work page

  13. [13]

    Experimental realization of any discrete unitary operator , author =. Phys. Rev. Lett. , volume =. 1994 , month =. doi:10.1103/PhysRevLett.73.58 , url =

    work page doi:10.1103/physrevlett.73.58 1994

  14. [14]

    2010 , publisher=

    Discrete-Time Signal Processing , author=. 2010 , publisher=

    work page 2010

  15. [15]

    Hernandez, J

    Vicente Hernandez and Jose E. Roman and Vicente Vidal. SLEPc : A scalable and flexible toolkit for the solution of eigenvalue problems. ACM Trans. Math. Software. 2005. doi:10.1145/1089014.1089019

    work page doi:10.1145/1089014.1089019 2005

  16. [16]

    Lehoucq, R. B. and Sorensen, D. C. and Yang, C. , title =. 1998 , address =

    work page 1998

  17. [17]

    L. S. Blackford and J. Choi and A. Cleary and E. D'Azevedo and J. Demmel and I. Dhillon and J. Dongarra and S. Hammarling and G. Henry and A. Petitet and K. Stanley and D. Walker and R. C. Whaley , title =. 1997 , address =

    work page 1997

  18. [18]

    Calcolo , volume =

    He, Haoze and Kressner, Daniel , title =. Calcolo , volume =. 2025 , doi =

    work page 2025