Weak supermajorization between symplectic spectra of positive definite matrix and its pinching

This paper establishes that for a $2n \times 2n$ real positive definite matrix, the vector of symplectic eigenvalues of its block diagonal part is weakly supermajorized by the vector of symplectic eigenvalues of the entire matrix, while also proving a related weak supermajorization inequality involving the eigenvalues of specific matrix products derived from the blocks.

The Gist

Imagine you have a giant, complex machine made of two main parts, let's call them Part E and Part G, connected by a messy, tangled web of wires (let's call this Part F). In the world of mathematics, this whole machine is a "Positive Definite Matrix." It's a system that is stable, balanced, and full of energy.

Now, every machine has a "heartbeat" or a "natural rhythm." In this specific field of math (symplectic geometry), this rhythm is called the Symplectic Spectrum. Think of it as the unique set of frequencies at which the machine vibrates.

The Big Idea: The "Pinching" Experiment

The paper asks a simple question: What happens to the machine's rhythm if we cut the wires?

Imagine taking a pair of scissors and snipping all the connections between Part E and Part G. You are left with two separate, isolated machines: just Part E and just Part G, sitting side-by-side but not talking to each other. In math, this is called a "Pinching."

The authors, Temjensangba and Mishra, wanted to know: Does the rhythm of the original, connected machine (A) dominate the rhythm of the two separated machines (E and G)?

The Answer: The "Supermajorization" Rule

Their discovery is a bit like a law of conservation of energy, but for rhythms. They found that the original, connected machine is always "stronger" or "more spread out" in its vibrations than the separated pieces.

They use a fancy term called Weak Supermajorization. Let's break that down with an analogy:

• The Separated Machines (E and G): Imagine you have two small buckets of water. If you pour them together, you get a total amount of water.
• The Connected Machine (A): Now imagine that same water, but it's been pressurized and mixed with a secret ingredient (the wires).

The paper proves that if you line up the "vibration strengths" of the separated machines from smallest to biggest, and do the same for the connected machine, the connected machine will always have more "oomph" at every step of the way.

The Metaphor:
Think of the separated machines as a group of people walking in a line, each carrying a small box.
Think of the connected machine as that same group, but now they are holding hands and walking in a synchronized, powerful dance.
The paper says: The synchronized dance (the connected machine) is always more powerful than the sum of the individuals walking alone. Even if you look at just the top 10% of the strongest walkers, the dance group will still be stronger.

Why is this cool? (The "Pinching" Surprise)

Usually, in math, when you cut things apart (pinching), you expect to lose information or power. This paper confirms that intuition but adds a twist:

1. The Original is King: The connected machine's rhythm always beats the separated pieces.
2. The "Diagonal" Trick: They also looked at what happens if you only keep the very center of the blocks (the diagonal) and throw away everything else. Even then, the original machine is still the strongest.

The "Hidden Gem" (A Side Discovery)

While proving this, they found a side result that is interesting on its own. They discovered a relationship between two different ways of calculating the "strength" of these machines.

Imagine you have two ways to measure the strength of a bridge:

1. Method A: Measure the bridge as it is built.
2. Method B: Measure the bridge after you've rearranged the steel beams in a specific, complex way.

They found that Method A is always "stronger" (in a specific mathematical sense) than Method B. This is a new rule for how these specific types of matrices behave, which didn't exist before.

The "What If" Warning (The Counter-Example)

The authors were careful. They showed that this rule only works if you cut the machine in a very specific way (keeping the two main blocks intact).

They gave an example where they cut the machine in a weird, random way (like cutting a slice out of the middle of the wire). In that case, the rule broke! The separated pieces actually became "stronger" than the original in some ways. This tells us that the "Pinching" has to be done carefully to keep the magic rule working.

Summary for the Everyday Person

• The Setup: You have a complex system made of two parts connected by wires.
• The Action: You cut the wires to separate the parts.
• The Result: The original, connected system is always "more powerful" in its vibrations than the separated parts.
• The Takeaway: Connection creates strength. When you take a complex, interconnected system and break it into isolated pieces, you lose a specific type of mathematical "power" or "spread" in its behavior.

This paper is a new rulebook for how these mathematical machines behave when we try to simplify them by cutting them apart. It helps scientists and engineers understand the limits of simplification in complex systems, from quantum physics to signal processing.

Technical Summary

1. Problem Statement

The paper addresses a gap in the theory of symplectic eigenvalues regarding their behavior under pinching operations.

• Context: Symplectic eigenvalues are fundamental in symplectic geometry and quantum information theory (e.g., Gaussian states). While classical eigenvalue theory has established majorization results for pinching (e.g., the eigenvalues of a Hermitian matrix majorize the eigenvalues of its pinching), the symplectic analogs are less developed.
• Specific Gap: Previous work (e.g., [12]) established that symplectic eigenvalues of a matrix weakly supermajorize the symplectic eigenvalues of its symplectic pinching. However, it was unknown whether a similar relation holds for the classical pinching of a positive definite matrix (specifically, retaining only the diagonal blocks of a block-partitioned matrix).
• Goal: The authors aim to establish a weak supermajorization relation between the symplectic spectrum of a $2n \times 2n$ real positive definite matrix $A$ and the symplectic spectrum of its classical pinching (the direct sum of its diagonal blocks, $E \oplus G$).

2. Methodology and Preliminaries

The authors utilize tools from linear algebra, symplectic geometry, and matrix analysis.

• Definitions:

  • Symplectic Eigenvalues: For a $2n \times 2n$ positive definite matrix $A$, Williamson's theorem guarantees the existence of a symplectic matrix $M$ such that $M^T A M = D \oplus D$, where $D$ is diagonal. The diagonal entries of $D$ are the symplectic eigenvalues, denoted $d(A)$.
  • Pinching: For a block matrix $A = \begin{bmatrix} E & F \\ F^T & G \end{bmatrix}$, the classical pinching $\mathcal{C}(A)$ relative to the block structure is defined as the direct sum of the diagonal blocks: $\mathcal{C}(A) = E \oplus G$.
  • Weak Supermajorization ($\prec_w$): A vector $x$ is weakly supermajorized by $y$ ($x \prec_w y$) if the sum of the $k$ smallest entries of $x$ is greater than or equal to the sum of the $k$ smallest entries of $y$ for all $k$.

• Key Lemma (Lemma 3.1):
  The authors establish a crucial identity linking the symplectic eigenvalues of a direct sum to the standard eigenvalues of a product of matrices:
  $$d(E \oplus G) = \lambda\left( (G^{1/2} E G^{1/2})^{1/2} \right)$$
  where $\lambda(\cdot)$ denotes the vector of standard eigenvalues in non-decreasing order. This connects symplectic spectral theory to standard spectral theory.

• Variational Characterization:
  The proof of the main theorem relies on constructing a specific symplectic basis for the matrix $E \oplus G$ and utilizing a variational principle for symplectic eigenvalues (referencing [12, Theorem 5]), which states that the sum of the first $k$ symplectic eigenvalues is the minimum trace of $X^T A X$ over specific symplectic subspaces.

3. Key Contributions and Results

Main Theorem (Theorem 3.2)

The central result of the paper is the weak supermajorization relation:
$$d(E \oplus G) \prec_w d(A)$$
Interpretation: The symplectic eigenvalues of the pinched matrix (the direct sum of the diagonal blocks) are weakly supermajorized by the symplectic eigenvalues of the original matrix $A$.

• Proof Sketch: The authors construct a specific symplectic matrix $M$ composed of eigenvectors of $E \oplus G$. They show that the trace of $M^T (E \oplus G) M$ equals the trace of $M^T A M$. By applying the variational characterization of symplectic eigenvalues, they demonstrate that the partial sums of the symplectic eigenvalues of $E \oplus G$ are bounded below by those of $A$.

Corollaries and Consequences

1. Corollary 3.4 (Connection to Standard Eigenvalues):
   Using Lemma 3.1, the authors derive:
   $$\lambda\left( (G^{1/2} E G^{1/2})^{1/2} \right) \prec_w d(A)$$
   This provides a bridge between the symplectic spectrum of $A$ and the standard eigenvalues of the geometric mean of its diagonal blocks.

2. Schur-type Analog (Equation 3.31):
   By combining the main result with known results on symplectic pinching of diagonal matrices, they establish:
   $$\Delta_s(A) \prec_w d(A)$$
   where $\Delta_s(A)$ contains the square roots of the products of the diagonal entries of $E$ and $G$. This is identified as a symplectic analog of the classic Schur's theorem.

3. Corollary 3.5 (Pinching of Diagonal Blocks):
   The authors show that pinching the diagonal blocks $E$ and $G$ further preserves the inequality:
   $$\lambda\left( (\mathcal{C}(G)^{1/2} \mathcal{C}(E) \mathcal{C}(G)^{1/2})^{1/2} \right) \prec_w \lambda\left( (G^{1/2} E G^{1/2})^{1/2} \right)$$
   This implies that the operation of pinching reduces the "magnitude" of the symplectic spectrum in the sense of weak supermajorization.

Counter-Example (Example 3.3)

The paper clarifies the scope of the result by showing that the weak supermajorization $d(\mathcal{C}(A)) \prec_w d(A)$ fails if the pinching is not performed on the specific $n \times n$ diagonal blocks (i.e., if the partitioning is arbitrary, such as $(1, 3)$ in a $4 \times 4$ matrix). This highlights the structural necessity of the $n \times n$ block partition for the theorem to hold.

4. Significance

• Theoretical Advancement: This work extends the classical theory of eigenvalue majorization (specifically Schur-Horn and Lidskii theorems) into the realm of symplectic geometry. It confirms that the "pinching" operation, a fundamental concept in matrix analysis, behaves predictably with respect to symplectic spectra under specific block structures.
• Quantum Information Relevance: Symplectic eigenvalues are directly related to the uncertainty principle and the entropy of Gaussian quantum states. Results regarding the ordering of these eigenvalues under operations like pinching (which can model partial tracing or coarse-graining in quantum systems) are vital for understanding entanglement and information flow in continuous variable quantum systems.
• New Inequalities: The paper provides new inequalities linking symplectic spectra to standard eigenvalues of matrix products (e.g., $G^{1/2} E G^{1/2}$), offering new tools for bounding symplectic capacities and entropies.

In summary, the paper successfully establishes that for a $2n \times 2n$ positive definite matrix partitioned into $n \times n$ blocks, the symplectic spectrum of the matrix is "larger" (in the sense of weak supermajorization) than the symplectic spectrum of its diagonal block direct sum, filling a significant gap in symplectic spectral theory.