---
title: Bivariate Flow Polynomial in Signed Graphs
url: https://www.emergentmind.com/topics/bivariate-flow-polynomial
type: topic
---

# Bivariate Flow Polynomial in Signed Graphs

Searching arXiv for the specified paper and closely related signed-graph/Tutte work mentioned in the source data.
{}
The **bivariate flow polynomial** of a signed graph is the unique polynomial \(F_\Sigma(x,y)\) such that, for every finite abelian group \(A\), the number of nowhere-zero \(A\)-flows on \(\Sigma\) is \(F_\Sigma(|A|,|A[2]|)\), where \(A[2]=\{a\in A:2a=0\}\) is the \(2\)-torsion subgroup. In the formulation developed for signed graphs with possible outer-edges, the invariant arises from a chain-group and homological framework in which flows are \(\ker \partial\), tensions are defined through orthogonality to integer-valued flows, and the frame matroid appears via minimal supports of nonzero flows. The second variable is forced by the signed setting: negative circles constrain flow values to \(A[2]\), so nowhere-zero flow counts are not governed solely by \(|A|\) [2508.11288].

## 1. Signed graphs, chain groups, and incidence structure

A signed graph is a triple \(\Sigma=(V,E,\sigma)\), where \(V\) is a finite vertex set, \(E\) is a finite edge set, and \(\sigma:E\to\{-1,+1\}\) assigns a sign to each edge. The framework allows **outer-edges** as well as loops and links. An outer-edge has one end attached to a vertex and the other end “outside”; the paper treats these edges as carrying a sign, while emphasizing that their role is topological through noncompactness and algebraic through incidence [2508.11288].

Each edge has two ends and admits a bi-directional orientation. If an edge is positive, the two arrows point in the same direction along the edge; if it is negative, they point in opposite directions. For a link or loop \(e=uv\), the orientations are:
- positive link or loop: \(u\to v\to\) and \(u\leftarrow v\leftarrow\),
- negative link or loop: \(u\leftarrow v\to\) and \(u\to v\leftarrow\).

For an outer-edge \(e=vo\), the orientations are:
- positive outer-edge: \(v\to o\to\) and \(v\leftarrow o\leftarrow\),
- negative outer-edge: \(v\leftarrow o\to\) and \(v\to o\leftarrow\).

The signed incidence function is defined on \(V\times \vec E\) by
\[
[v,\vec e]=
\begin{cases}
1 & \text{if } \vec e \text{ is a non-loop arc pointing to } v,\\
2 & \text{if } \vec e \text{ is a negative loop arc pointing to } v,\\
0 & \text{otherwise,}
\end{cases}
\]
and extended by \([v,\vec e^{-1}]=-[v,\vec e]\). After fixing an orientation \(\omega\subseteq \vec E\), one obtains the signed incidence matrix \(M=[m_{ve}]\) with \(m_{ve}=[v,\vec e]\) for \(\vec e\in\omega\) [2508.11288].

With coefficients in an abelian group \(A\), the \(1\)-chain group \(C_1(\Sigma;A)\) consists of functions \(c:\vec E\to A\) satisfying \(c(\vec e^{-1})=-c(\vec e)\), and the \(0\)-chain group \(C_0(\Sigma;A)\) consists of functions \(p:V\to A\). The boundary operator
\[
\partial:C_1(\Sigma;A)\to C_0(\Sigma;A),\qquad (\partial c)(v)=\sum_{e\in E}[v,\vec e]\,c(\vec e),
\]
has the equivalent edgewise description
- \(\partial \vec e=v-u\) for a positive link or loop \(u\to v\to\),
- \(\partial \vec e=u+v\) for a negative link or loop \(u\leftarrow v\to\),
- \(\partial \vec e=v\) for an outer-edge arc \(v\leftarrow o\sim\).

The coboundary operator
\[
\delta:C_0(\Sigma;A)\to C_1(\Sigma;A),\qquad \delta p(\vec e)=\sum_{v\in V}[v,\vec e]\,p(v),
\]
is equivalently given by
- \(\delta p(\vec e)=p(v)-p(u)\) for \(u\to v\to\),
- \(\delta p(\vec e)=p(u)+p(v)\) for \(u\leftarrow v\to\),
- \(\delta p(\vec e)=p(u)\) for \(u\leftarrow o\sim\),
with \(\delta p(\vec e^{-1})=-\delta p(\vec e)\) [2508.11288].

This setup standardizes how the sign of an edge affects conservation. In particular, negative edges replace differences by sums, and negative loops contribute \(2\) at their incident vertex.

## 2. Flows, tensions, and homological structure

The **flow group** is
\[
F(\Sigma;A)=\ker \partial \subseteq C_1(\Sigma;A).
\]
A nowhere-zero \(A\)-flow is a flow \(f\in F(\Sigma;A)\) such that \(f(\vec e)\neq 0\) for all edges \(e\). The dependence on signed incidence is explicit: on negative edges the conservation law uses \(u+v\) rather than \(v-u\), and on a negative loop the local contribution is doubled [2508.11288].

The **tension group** is defined through the canonical bilinear pairing
\[
\langle \cdot,\cdot\rangle:C_i(\Sigma;\mathbb Z)\times C_i(\Sigma;A)\to A,\qquad i=0,1,
\]
where
\[
\langle c,d\rangle=\sum_{v\in V} c(v)d(v)\quad\text{if } i=0,
\qquad
\langle c,d\rangle=\sum_{e\in E} c(\vec e)d(\vec e)\quad\text{if } i=1.
\]
An element \(g\in C_1(\Sigma;A)\) is a tension if \(\langle f,g\rangle=0\) for all \(f\in F(\Sigma;\mathbb Z)\). The paper states that \(\operatorname{im}\delta\subseteq T(\Sigma;A)\) and that the adjoint identity
\[
\langle \partial c,p\rangle=\langle c,\delta p\rangle
\]
holds [2508.11288].

The boundary group and the homology and cohomology groups are
\[
B_0(\Sigma;A)=\operatorname{im}\partial,\qquad
H_0(\Sigma;A)=C_0(\Sigma;A)/B_0(\Sigma;A),\qquad
H^1(\Sigma;A)=C_1(\Sigma;A)/\operatorname{im}\delta.
\]
A central structural point is that this chain-group formalism is not merely notational. It aligns the combinatorics of signed graphs with algebraic-topological operators and makes flows, tensions, homology, and cohomology available uniformly even in the presence of outer-edges. This suggests a deliberate replacement of ad hoc signed-graph case analysis by a systematic linear-algebraic and homological treatment.

## 3. Frame matroid interpretation

The paper’s guiding principle is the correspondence between representable matroids over \(\mathbb R\) on a ground set \(E\) and subspaces of the vector space of real-valued chains on the same ground set. In this setting, the frame matroid of a signed graph emerges by defining circuits as minimal supports of nonzero flows rather than by listing circuit patterns independently [2508.11288].

Let \(F(\Sigma;\mathbb R)\) be the real-valued flow space, with the standard inner product on \(C_1(\Sigma;\mathbb R)\). An **elementary flow** is a nonzero flow whose support is minimal under inclusion. The supports of elementary flows form the circuit family of a matroid \(M(F)\) on \(E\), and this matroid is the frame matroid \(M(\Sigma)\). The signed-graph circuits are precisely the following five types:

- **(C1)** Open line \(L\).
- **(C2)** Positive circle \(C\).
- **(C3)** Two negative circles sharing a unique common vertex.
- **(C4)** A negative circle together with a half-line meeting at the endpoint of the half-line.
- **(C5)** Two disjoint negative circles connected by a line segment.

For a chosen direction \(\omega(C)\) of such a circuit, the principal integer-valued circular flow is defined by
\[
I_{\omega(C)}(\vec e)=
\begin{cases}
1 & \text{if } \vec e\in \omega(C)\text{ and } e\text{ is single},\\
2 & \text{if } \vec e\in \omega(C)\text{ and } e\text{ is doubled},\\
0 & \text{otherwise.}
\end{cases}
\]

Dually, tensions span a subspace \(T(\Sigma;\mathbb R)=F(\Sigma;\mathbb R)^\perp\), and elementary tensions yield the circuits of the dual matroid \(M^*(\Sigma)\). These cocircuits are precisely the **bonds** of the signed graph: minimal edge sets \(U\subseteq E\) such that
\[
b(\Sigma\setminus U)=b(\Sigma)+1
\]
and
\[
b((\Sigma\setminus U)\cup e)=b(\Sigma)\quad\text{for every } e\in U.
\]
Equivalently, a bond is a minimal support of a nonzero tension [2508.11288].

Deletion and contraction follow signed-graphic edge operations with special care for negative loops and outer-edges. Positive-link contraction merges endpoints; a bridge or an outer-edge co-loop yields no additional flows; a positive loop contributes a direct \(A\)-summand to \(F\); and a negative loop that is also a co-loop contributes an \(A[2]\)-summand. This is the matroidal mechanism behind the later recurrence formulas.

## 4. Definition, expansion, and the necessity of two variables

For a signed graph \(\Sigma\), the bivariate flow polynomial \(F_\Sigma(x,y)\) is characterized by
\[
|F_{\mathrm{nz}}(\Sigma;A)|=F_\Sigma(|A|,|A[2]|)
\]
for every finite abelian group \(A\). The paper also uses \(\phi(\Sigma;t,x)\) with \(t=|A[2]|\) and \(x=|A|\), and sets \(F_\Sigma(x,y):=\phi(\Sigma;y,x)\) [2508.11288].

Its main expansion theorem is
\[
F_\Sigma(x,y)=\sum_{S\subseteq E}(-1)^{|E|-|S|}y^{u_c(S)}x^{cr(S)},
\]
where \(u_c(S)\) is the number of compact unbalanced components of the spanning subgraph \((V,S)\), and \(cr(S)\) is its cycle rank, equivalently the matroid corank
\[
r^*(S)=|S|-r(S),\qquad r(S)=|V|-b(S).
\]
The theorem implies that the number of nowhere-zero \(A\)-flows depends only on \(|A|\) and \(|A[2]|\) [2508.11288].

The reason a univariate invariant fails in general is explicit. For unsigned or balanced graphs, nowhere-zero flow counts depend only on \(|A|\), producing the ordinary univariate flow polynomial. In signed graphs, however, every flow on a directed negative circle \(C\) has the form \(a\cdot 1_{\omega(C)}\) with \(a\in A[2]\). Thus the cardinality of the \(2\)-torsion subgroup directly affects enumeration. For \(A=\mathbb Z_n\), one has \(|A[2]|=1\) when \(n\) is odd and \(|A[2]|=2\) when \(n\) is even, so the counting functions differ by parity class [2508.11288].

The simplest examples already exhibit the dichotomy:
- a single negative loop has \(F_\Sigma(x,y)=y-1\),
- a single positive loop has \(F_\Sigma(x,y)=x-1\),
- a single positive link has \(F_\Sigma(x,y)=0\).

A plausible implication is that the second variable \(y\) is not a refinement added for formal symmetry, but the minimal additional parameter needed to record the obstruction created by negative cycles.

## 5. Structural properties, recurrences, and degrees

The polynomial satisfies deletion–contraction recurrences that mirror the edge taxonomy of signed graphs. If \(e\in E\), then [2508.11288]:
- if \(e\) is a positive loop,
  \[
  F_\Sigma(x,y)=(x-1)F_{\Sigma-e}(x,y);
  \]
- if \(e\) is a negative loop and also a co-loop,
  \[
  F_\Sigma(x,y)=(y-1)F_{\Sigma-e}(x,y);
  \]
- if \(e\) is a positive link,
  \[
  F_\Sigma(x,y)=F_{\Sigma/e}(x,y)-F_{\Sigma-e}(x,y).
  \]

The same source states that \(F_\Sigma(x,y)=0\) if and only if \(\Sigma\) contains a co-loop that is either a bridge co-loop or an outer-edge co-loop. The base cases are:
- edgeless signed graph: \(F_\Sigma(x,y)\equiv 1\),
- single positive loop: \(x-1\),
- single negative loop: \(y-1\),
- single outer-edge: \(0\),
- single positive link: \(0\).

Two invariance properties are singled out. First, \(F_\Sigma\) is multiplicative on disjoint unions:
\[
F_\Sigma(x,y)=\prod_i F_{\Sigma_i}(x,y)
\]
when \(\Sigma=\bigsqcup_i \Sigma_i\). Second, it is switching invariant, because switching preserves flows, tensions, and the signed incidence data up to isomorphism [2508.11288].

The paper also ties polynomial degrees to signed-graphic rank parameters. Writing
\[
F_\Sigma(x,y)=\sum_{i=0}^d y^i F_i(x),
\]
one has \(d\le cr^-(\Sigma)\), where \(cr^-(\Sigma)\) is the negative cycle-rank. More specifically:
- if \(\Sigma\) has no negative circles, then \(F_\Sigma(x,y)=F_0(x)\) is monic of degree \(cr(\Sigma)\);
- if \(\Sigma\) has outer-edges, then \(F_0(x)\) is monic of degree \(cr(\Sigma)\);
- if \(\Sigma\) has negative circles and no outer-edges, then \(\deg F_0(x)\le cr^+(\Sigma)\) and \(F_0\) may vanish;
- if \(\Sigma\) has negative circles and no outer-edges, then \(F_1(x)\) is monic of degree \(cr(\Sigma)\) [2508.11288].

These statements place the \(x\)- and \(y\)-degrees under matroidal control, while separating the role of total cycle rank from the contribution of negative cycles.

## 6. Classical specializations, coefficient interpretations, and examples

When \(\Sigma\) is balanced and contains no outer-edges, the bivariate polynomial specializes to the ordinary graph flow polynomial:
\[
F_\Sigma(x,y)=\sum_{S\subseteq E}(-1)^{|E|-|S|}x^{cr(S)}=:\phi_G(x).
\]
In this case \(u_c(S)\equiv 0\), and the signed theory collapses to the unsigned one. In particular, for cycles \(C_n\), \(\phi_{C_n}(x)=x-1\), and for general graphs \(G\), \(\phi_G(k)\) counts nowhere-zero \(\mathbb Z_k\)-flows [2508.11288].

The relation to tensions and Tutte-type invariants is stated in dual language. Tensions are dual to flows through the orthogonal decomposition \(C_1(\Sigma;\mathbb R)=F\oplus T\). For balanced graphs, the tension polynomial is dual to the flow polynomial through the Tutte polynomial \(T_G\). For signed graphs and the frame matroid \(M(\Sigma)\), \(F_\Sigma(x,y)\) is described as a switching-invariant bivariate invariant and as an inclusion–exclusion evaluation weighted simultaneously by cycle rank and unbalanced compact components. The paper notes that a signed-graphic dichromate due to Goodall–Litjens–Regts–Vena extends Tutte’s universality, while a full Tutte-type universality for \(F_\Sigma\) itself requires multivariate parameters for balanced and unbalanced components [2508.11288].

The coefficient interpretation follows directly from the expansion:
\[
F_\Sigma(x,y)=\sum_{S\subseteq E}(-1)^{|E|-|S|}y^{u_c(S)}x^{cr(S)}.
\]
Thus the coefficient of \(y^i x^j\) is the alternating sum over subsets \(S\) with \(u_c(S)=i\) and \(cr(S)=j\). Here \(cr(S)\) counts independent cycles, while \(u_c(S)\) counts unbalanced compact components. The maximal exponent of \(y\) equals \(cr^-(\Sigma)\), described as the negative cycle-rank [2508.11288].

Several concrete examples are given:
- a single negative link between two vertices has \(F_\Sigma(x,y)=0\);
- a balanced cycle \(C_n\) has \(F_{C_n}(x,y)=x-1\);
- an unbalanced cycle with one negative edge has \(F_\Sigma(x,y)=y-1\);
- two loops at a vertex, one positive and one negative, give
  \[
  F=(x-1)(y-1);
  \]
- a triangle with all positive edges gives
  \[
  F=(x-1)^2;
  \]
- a triangle with one negative edge gives
  \[
  F=(x-1)(x-y).
  \]

For the family \(\Sigma_{m,n}\) with one vertex, \(m\) outer-edges, and \(n\) negative loops, the paper records the base values
\[
F_{1,0}=0,\qquad F_{0,1}=y-1,\qquad F_{1,1}=x-y,
\]
the recurrences
\[
F_{0,n}=y(x-1)^{n-1}-F_{0,n-1}\quad (n\ge 1),
\]
\[
F_{m,n}=(x-1)^{m+n-1}-F_{m-1,n}\quad (m\ge 1),
\]
and the closed forms
\[
F_{m,0}=\frac{(x-1)^m-(-1)^m}{x}+(-1)^m,
\]
\[
F_{0,n}=y\frac{(x-1)^n-(-1)^n}{x}+(-1)^n,
\]
\[
F_{m,n}=(x-1)^n\frac{(x-1)^m-(-1)^m}{x}+(-1)^m y\frac{(x-1)^n-(-1)^n}{x}+(-1)^{m+n}.
\]

The paper also describes a parity phenomenon. For \(A=\mathbb Z_n\), one has
\[
F_\Sigma(n)=F_\Sigma(n,2)\quad\text{for even } n,
\qquad
F_\Sigma(n)=F_\Sigma(n,1)\quad\text{for odd } n.
\]
Accordingly, the counting function in \(n\) is a quasi-polynomial of degree \(cr(\Sigma)\) with period \(2\) when negative cycles are present, and a genuine polynomial when \(\Sigma\) is balanced [2508.11288].

## 7. Computation, complexity, and broader consequences

Two general computation methods are identified. The first is the inclusion–exclusion expansion over all \(S\subseteq E\), which requires determining \(u_c(S)\) and \(cr(S)\) for each subset and is therefore exponential in \(|E|\). The second is deletion–contraction based on the recurrences for positive loops, negative loop co-loops, and positive links, again exponential in general [2508.11288].

The paper further states that dynamic programming yields polynomial-time computation on special classes, specifically cactus graphs, series–parallel signed graphs, and graphs with bounded treewidth, because on those subclasses \(u_c\) and \(cr\) can be computed efficiently. It also notes that computing \(F_\Sigma(x,y)\) generalizes computation of the Tutte polynomial and of the ordinary flow polynomial for unsigned graphs, and is therefore expected to be \(\#\mathrm P\)-hard in general [2508.11288].

The broader significance attributed to the invariant is structural. The chain-group framework introduces boundary and coboundary operators intrinsically for signed graphs with outer-edges, harmonizing flows and tensions with algebraic topology and matroid theory. Within this framework:
- flows are \(\ker \partial\),
- tensions are \(\operatorname{im}\delta\) and, more generally, orthogonal complements annihilating integer flows,
- circuits and bonds emerge as minimal supports of elementary flows and tensions,
- the frame matroid is identified through elementary chains,
- the bivariate flow polynomial resolves the “mystery” of the nonexistence of a univariate flow polynomial for signed graphs by isolating the essential dependence on \(|A[2]|\) [2508.11288].

In that sense, the bivariate flow polynomial occupies the role that the ordinary flow polynomial plays for balanced graphs, but in a form adapted to signed incidence, negative cycles, and outer-edges. It is simultaneously a counting invariant, a deletion–contraction invariant, and a matroidally interpretable invariant derived from homological data.

Source: https://www.emergentmind.com/topics/bivariate-flow-polynomial