## Rational Formulas for Traces in zero-dimensional Algebras

### Gabriela JeronimoPartially supported by the Argentinian research grants UBACyT X112 (2004-2007) and CONICET PIP 02461/01. * * Corresponding author.

Abstract
Let $\mathbb{K}$  be a field of characteristic zero, and $S:=\mathbb{K}\left[{x}_{1},\dots ,{x}_{n}\right].$  If $\mathcal{ℐ}\subset S$  is a zero-dimensional ideal and $r\in \mathbb{K}\left({x}_{1},...,{x}_{n}\right)$  a rational function whose denominator is not a zero divisor in $A:=S/\mathcal{ℐ},$  we establish a rational expression for the trace of the multiplication map ${\text{Times}}_{r}:A\to A$  in terms of the generalized Chow form of $\mathcal{ℐ}.$  In the case $\mathcal{ℐ}$  is a complete intersection in the torus, we get numerator and denominator formulas for traces in terms of sparse resultants.
For complete intersection in the affine space without zeroes at the infinity, we recover classical results about multidimensional residues with our methods.

1 Introduction

Traces in finite dimensional algebras play a fundamental role in Commutative Algebra and Algebraic Geometry. Recent applications of traces include the evaluation of symmetric functions, the effective Nullstellensatz and algorithms for solving polynomial systems [ABRW, BW, DG, FGS, SS, Kun.
For developments in the analytical counterpart of this algebraic tool (residues) see [AGV, AY, CDS1, CDS2, CM, Tsi.
Let $\mathbb{K}$  be a field of characteristic zero, and $S:=\mathbb{K}\left[{x}_{1},\dots ,{x}_{n}\right].$  Let $\mathcal{ℐ}$  be a zero-dimensional ideal of $S,$  and set $A:=S/\mathcal{ℐ}.$  Let $p,q\in S$  such that $q$  is not a zero divisor in $A.$  Then, $r:=\frac{p}{q}$  induces a $\mathbb{K}$  -linear map
 $\begin{array}{c}\begin{array}{cccc}{\text{Times}}_{r}:& A& \to & A\\ & a& ↦& r\cdot a.\end{array}\end{array}$ (1)
We are interested in the computation of the trace of this map. First, we relate $Trace\left({\text{Times}}_{r}\right)$  with an algebraic object depending on $\mathcal{ℐ}$  and $r$  : the generalized Chow form of the ideal $\mathcal{ℐ}$  (see [Phifor the definition of this eliminating polynomial).
Then, we focus on the case where $\mathcal{ℐ}$  is given by a complete intersection in the torus. We show that, in this case, the generalized Chow form may be replaced with a sparse resultant (in the sense of [CLO, GKZ), and we exhibit numerator and denominator formulas for traces in terms of sparse resultants, similar to the denominator formulas obtained for residues in [CDS2. In particular, our trace formulas can be used for computing global residues.
Finally, we compare our formulas for the denominator of the trace with those obtained from the formulas proposed in [CDS2for denominators of residues in the torus.
Effective procedures for the computation of traces can be derived from our formulas.
In order to do this, one has to deal with Chow forms and resultants. In the case $\mathcal{ℐ}$  is a radical ideal, the generalized Chow form of $\mathcal{ℐ}$  coincides with the generalized Chow form of the variety $V\left(\mathcal{ℐ}\right)$  and there are effective algorithms for its computation (see, for instance, [JKSS). Algorithms for the computation of sparse resultants can be found in [CLO,Chapter7.
The paper is organized as follows: In Section  2 we show a general formula for the computation of the trace based on the generalized Chow form. In Section  3 , we focus on the case where we have a generic complete intersection in the torus. We give rational expressions for both the numerator and the denominator of the trace. We compare our results with those obtained in [CDS2for the computation of residues in Section  4 . Section  5 is concerned with multidimensional residues on affine space. By using results of Jouanolou we can recover denominator formulas for residues given in [Tsi, CDS1, CDS2and also give an algebraic proof of the Euler-Jacobi formula.
Acknowledgements. We are grateful to Jean Pierre Jouanolou for helpful conversations about his work on discriminants and also to Abdallah Al-Amrani for sharing with us the notes of the course [Jou.

2 A rational expression for the trace

As in the introduction, let $\mathbb{K}$  be a field of characteristic zero, and $S:=\mathbb{K}\left[{x}_{1},\dots ,{x}_{n}\right].$  Let $\mathcal{ℐ}$  be a zero-dimensional ideal of $S,$  and set $A:=S/\mathcal{ℐ}.$  Let $p,q\in S$  such that $q$  is not a zero divisor in $A.$  Set $d:=max\left\{deg\left(p\right),deg\left(q\right)\right\},$  where $deg$  denotes total degree in the variables ${x}_{1},\dots ,{x}_{n}.$  We introduce new variables ${U}_{\alpha }$  for $\alpha \in \left({\mathbb{Z}}_{\ge 0}{\right)}^{n},$  $|\alpha |\le d.$  Let $U:=\left({U}_{\alpha },|\alpha |\le d\right)$  and $\mathcal{C}{h}_{d,\mathcal{ℐ}}\left(U\right)$  be the generalized Chow form of $\mathcal{ℐ}$  (see [Phi): if $\mathcal{U}\left(x\right):={\sum }_{|\alpha |\le d}{U}_{\alpha }{x}^{\alpha },$  then $\mathcal{C}{h}_{d,\mathcal{ℐ}}\left(U\right)={\prod }_{\xi \in V\left(\mathcal{ℐ}\right)}\mathcal{U}\left(\xi {\right)}^{m\left(\xi \right)},$  where $m\left(\xi \right)$  is the multiplicity of $\xi$  with respect to $\mathcal{ℐ},$  i.e. the dimension of the local ring ${A}_{\xi }:={S}_{\xi }/\mathcal{ℐ}{S}_{\xi }.$
Theorem 2.1. Let $Time{s}_{r}$  be the map defined in ( 1 ). Then,
 $\begin{array}{c}Trace\left(Time{s}_{r}\right)=\frac{{\sum }_{|\alpha |\le d}{p}_{\alpha }\frac{\partial \mathcal{C}{h}_{d,\mathcal{ℐ}}}{\partial {U}_{\alpha }}\left(q\right)}{\mathcal{C}{h}_{d,\mathcal{ℐ}}\left(q\right)}\end{array}$ (2)
Observe that as $q$  is not a zero divisor in $A,$  then the denominator of ( 2 ) is not zero.
• Proof. Let $T$  be a new variable and set $\mathcal{X}\left(T\right):=\mathcal{C}{h}_{d,\mathcal{ℐ}}\left(q+Tp\right)={C}_{0}+{C}_{1}T+\text{higher terms in}T.$  Using the identity $\mathcal{C}{h}_{d,\mathcal{ℐ}}\left(q+Tp\right)={\prod }_{\xi \in V\left(\mathcal{ℐ}\right)}\left(q\left(\xi \right)+Tp\left(\xi \right){\right)}^{m\left(\xi \right)},$  it is easy to see that $Trace\left({\text{Times}}_{r}\right)$  equals $\frac{{C}_{1}}{{C}_{0}}=\frac{{\mathcal{X}}^{\prime }\left(0\right)}{\mathcal{X}\left(0\right)}.$  By applying the chain rule to compute the numerator, we get Identity ( 2 ).

3 Traces in the torus

Now we turn our attention to generic complete intersections in the torus. We will give rational formulas for traces of linear maps in terms of sparse resultants.

3.1 Rational formulas for traces in the torus

Consider a system of $k$  generic Laurent polynomials ${f}_{i}:={\sum }_{a\in {\mathcal{A}}_{i}}{c}_{ia}{\mathbf{t}}^{a}$  , where $i=1,\dots ,k,$  ${\mathcal{A}}_{i}$  is a finite subset of ${\mathbb{Z}}^{k}$  , the coefficients ${c}_{ia}$  are indeterminates over $\mathbb{Q},$  $a=\left({a}_{1},\dots ,{a}_{k}\right)\in {\mathbb{Z}}^{k}$  and ${\mathbf{t}}^{a}={{t}_{1}}^{{a}_{1}}{{t}_{2}}^{{a}_{2}}\dots {{t}_{k}}^{{a}_{k}}.$  For $i=1,...,k$  , the support of ${f}_{i}$  is the set of exponent vectors ${\mathcal{A}}_{i}\subset {\mathbb{Z}}^{k},$  and its Newton polytope is the convex hull ${P}_{i}=conv\left({\mathcal{A}}_{i}\right)\subset {\mathbb{R}}^{k}.$  We assume that the lattice affinely generated by ${\mathcal{A}}_{1},\dots ,{\mathcal{A}}_{k}$  , which we will denote with $L\left({\mathcal{A}}_{1},\dots ,{\mathcal{A}}_{k}\right),$  is a $k$  -dimensional affine sublattice of ${\mathbb{Z}}^{k}.$  Let $M$  be the mixed volume of the sequence of polytopes ${P}_{1},\dots ,{P}_{k}$  (see [CLOfor a definition).
Let $K$  be an algebraic closure of $\mathbb{Q}\left({c}_{ia}{\right)}_{1\le i\le k,a\in {\mathcal{A}}_{i}}$  and $S:=K\left[{t}_{1},{{t}_{1}}^{-1},\dots ,{t}_{k},{{t}_{k}}^{-1}\right].$  Then $A:=S/〈{f}_{1},\dots ,{f}_{k}〉$  is a finite-dimensional $K$  -vector space (see [PS). Let $\mathcal{A},{\mathcal{A}}^{\prime }$  be finite subsets of ${\mathbb{Z}}^{k}$  and consider a generic rational function of the form $r\left(\mathbf{t}\right):=\frac{p\left(\mathbf{t}\right)}{q\left(\mathbf{t}\right)}$  where $p\left(\mathbf{t}\right)={\sum }_{a\in \mathcal{A}}{p}_{a}{\mathbf{t}}^{a},q\left(\mathbf{t}\right)={\sum }_{a\in {\mathcal{A}}^{\prime }}{q}_{a}{\mathbf{t}}^{a},$  and ${p}_{a},{q}_{a}$  are new indeterminates. As $q\left(\mathbf{t}\right)$  is a generic denominator, it is invertible in $A.$  We can then consider the $K$  -linear map ${\text{Times}}_{r}:A\to A$  as defined in ( 1 ).
Let ${\mathcal{A}}_{0}:=\mathcal{A}\cup {\mathcal{A}}^{\prime }$  and ${f}_{0}={\sum }_{a\in {\mathcal{A}}_{0}}{c}_{0a}{\mathbf{t}}^{a},$  where ${c}_{0a}$  are new variables. As in [Stu, for $I\subset \left\{0,1,...,k\right\}$  we denote with ${L}_{I}$  the lattice affinely generated by ${\mathcal{A}}_{i},i\in I.$  A collection of supports $\left\{{\mathcal{A}}_{i}{\right\}}_{i\in I}$  is said to be essential if $\text{rank}\left({L}_{I}\right)=#I-1$  and $\text{rank}\left({L}_{J}\right)\ge #J$  for each proper subset $J$  of $I.$  From now on, we will assume that $M>0$  ; otherwise, our problem has no interest. Then, ${\mathcal{A}}_{0},{\mathcal{A}}_{1},\dots ,{\mathcal{A}}_{k}$  has a unique essential subset containing ${\mathcal{A}}_{0}.$  Moreover, the sparse resultant operator $Re{s}_{{\mathcal{A}}_{0},\dots ,{\mathcal{A}}_{k}}$  as defined in [Stuis not constantly one.
For a family of sets ${\mathcal{ℬ}}_{0},\dots ,{\mathcal{ℬ}}_{s}\subset {\mathbb{Z}}^{k},$  if $\left\{{\mathcal{ℬ}}_{i}{\right\}}_{i\in I}$  is the unique essential subfamily, write $L\left({\mathcal{ℬ}}_{0},\dots ,{\mathcal{ℬ}}_{s}\right)=sat\left(L\left({\mathcal{ℬ}}_{i},i\in I\right)\right)\oplus {L}^{\perp },$  where $sat\left(\right)$  denotes saturation with respect to the ambient lattice, and ${L}^{\perp }$  is the orthogonal complement. If ${L}^{\perp }=0,$  we define $\mathbf{e}\mathcal{ℬ}0,\dots ,\mathcal{ℬ}s:=1.$  Otherwise, denote with $\pi$  the projection onto the second factor and define $\mathbf{e}\mathcal{ℬ}0,\dots ,\mathcal{ℬ}s$  as the normalized mixed volume of the family $\left\{\pi \left(conv\left({\mathcal{ℬ}}_{i}\right)\right){\right\}}_{i/\in I}$  in ${L}^{\perp }$  ([Min).
One can use $Re{s}_{{\mathcal{A}}_{0},\dots ,{\mathcal{A}}_{k}}$  instead of the generalized Chow form used in the previous section, and the analogue of Theorem  2.1 is the following:
Theorem 3.1. Under the previous assumptions and notations,
 $\begin{array}{c}Trace\left(Time{s}_{r}\right)=\mathbf{d}\mathcal{A}0,\dots ,\mathcal{A}k\frac{{\sum }_{a\in {\mathcal{A}}_{0}}{p}_{a}\frac{\partial Re{s}_{{\mathcal{A}}_{0},\dots ,{\mathcal{A}}_{k}}}{\partial {c}_{0a}}\left(q,{f}_{1},\dots ,{f}_{k}\right)}{Re{s}_{{\mathcal{A}}_{0},\dots ,{\mathcal{A}}_{k}}\left(q,{f}_{1},\dots ,{f}_{k}\right)},\end{array}$ (3)
where $\mathbf{d}\mathcal{A}0,\dots ,\mathcal{A}k:=\left[{\mathbb{Z}}^{k}:L\left({\mathcal{A}}_{0},\dots ,{\mathcal{A}}_{k}\right)\right]\mathbf{e}\mathcal{A}0,\dots ,\mathcal{A}k.$
• Proof. Suppose first that $\left[{\mathbb{Z}}^{k}:L\left({\mathcal{A}}_{0},\dots ,{\mathcal{A}}_{k}\right)\right]=1.$  By using [Min,Lemma 13 , we have that $Re{s}_{{\mathcal{A}}_{0},\dots ,{\mathcal{A}}_{k}}\left({f}_{0},{f}_{1},\dots ,{f}_{k}{\right)}^{\mathbf{e}\mathcal{A}0,\dots ,\mathcal{A}k}=C{\prod }_{\gamma \in V\left({f}_{1},\dots ,{f}_{k}\right)}{f}_{0}\left(\gamma \right),$  with $C\in {K}^{*}.$
Remark 3.2. The reader should be cautious when comparing our results with Minimair's results in [Min. Indeed, in Remark $3$  of [Minthe resultant is defined as a power of an irreducible polynomial. In this paper, a resultant is always an irreducible polynomial.
Now, as in the proof of Theorem  2.1 , we make the substitution ${f}_{0}↦q+Tp,$  and get  $\begin{array}{c}Re{s}_{{\mathcal{A}}_{0},\dots ,{\mathcal{A}}_{k}}\left(q+Tp,{f}_{1},\dots ,{f}_{k}{\right)}^{\mathbf{e}\mathcal{A}0,\dots ,\mathcal{A}k}=\end{array}$
 $\begin{array}{c}Re{s}_{{\mathcal{A}}_{0},\dots ,{\mathcal{A}}_{k}}\left(q,{f}_{1},\dots ,{f}_{k}{\right)}^{\mathbf{e}\mathcal{A}0,\dots ,\mathcal{A}k}+\end{array}$
 $\begin{array}{c}+\mathbf{e}\mathcal{A}0,\dots ,\mathcal{A}kRe{s}_{{\mathcal{A}}_{0},\dots ,{\mathcal{A}}_{k}}\left(q,{f}_{1},\dots ,{f}_{k}{\right)}^{\mathbf{e}\mathcal{A}0,\dots ,\mathcal{A}k-1}\frac{\partial Re{s}_{{\mathcal{A}}_{0},\dots ,{\mathcal{A}}_{k}}\left(q+Tp,{f}_{1},\dots ,{f}_{k}\right)}{\partial T}{|}_{T=0}T+\end{array}$
 $\begin{array}{c}+\text{higher terms in}T.\end{array}$
By using the chain rule, the claim holds.
In the general case, [Min,Lemma 13 reads as follows:
$Re{s}_{{\mathcal{A}}_{0},\dots ,{\mathcal{A}}_{k}}\left({f}_{0},{f}_{1},\dots ,{f}_{k}{\right)}^{\mathbf{e}\mathcal{A}0,\dots ,\mathcal{A}k}=C{\prod }_{\beta \in {V}^{*}\left({f}_{1},\dots ,{f}_{k}\right)}{f}_{0}\left(\beta \right),$  with $C\in {K}^{*},$  and ${V}^{*}$  is the set of zeros with respect to the lattice generated by ${\mathcal{A}}_{0},\dots ,{\mathcal{A}}_{k}.$  The claim will hold if we show that  $\begin{array}{c}{\prod }_{\beta \in {V}^{*}\left({f}_{1},\dots ,{f}_{k}\right)}{f}_{0}\left(\beta {\right)}^{\left[{\mathbb{Z}}^{k}:L\left({\mathcal{A}}_{0},\dots ,{\mathcal{A}}_{k}\right)\right]}={\prod }_{\gamma \in V\left({f}_{1},\dots ,{f}_{k}\right)}{f}_{0}\left(\gamma \right).\end{array}$ (4)
By using normal Smith form reduction (this change preserves resultants, see [Min), we can suppose that $L\left({\mathcal{A}}_{0},\dots ,{\mathcal{A}}_{k}\right)={d}_{1}\mathbb{Z}\oplus \dots \oplus {d}_{k}\mathbb{Z},$  with $\left[{\mathbb{Z}}^{k}:L\left({\mathcal{A}}_{0},\dots ,{\mathcal{A}}_{k}\right)\right]={d}_{1}{d}_{2}\dots {d}_{k}.$  Then, ( 4 ) follows straightforwardly from [Min,Corollary 5 .
Example 3.3. Consider the following trivariate system:
$\left\{\begin{array}{ccc}{f}_{1}& =& {c}_{11}+{c}_{12}{t}_{1}^{2}+{c}_{13}{t}_{2}^{2}\\ {f}_{2}& =& {c}_{21}+{c}_{22}{t}_{2}^{2}+{c}_{23}{t}_{3}^{2}\\ {f}_{3}& =& {c}_{31}+{c}_{32}{t}_{1}^{2}+{c}_{33}{t}_{2}^{2}+{c}_{34}{t}_{3}^{2},\end{array}$  and let $\mathcal{A}=\left\{\left(0,2,0\right)\right\},{\mathcal{A}}^{\prime }:=\left\{\left(0,0,0\right)\right\}.$  In this case, the family $\left\{{\mathcal{A}}_{0},{\mathcal{A}}_{1},{\mathcal{A}}_{2},{\mathcal{A}}_{3}\right\}$  is essential, so ${\mathbf{e}}_{{\mathcal{A}}_{0},\dots ,{\mathcal{A}}_{3}}=1$  , but $\left[{\mathbb{Z}}^{3}:L\left({\mathcal{A}}_{0},\dots ,{\mathcal{A}}_{3}\right)\right]=8.$  A straightforward computation shows that $Trace\left({\text{Times}}_{{t}_{2}^{2}}\right)=8\frac{{c}_{11}{c}_{32}{c}_{23}+{c}_{21}{c}_{12}{c}_{34}-{c}_{31}{c}_{12}{c}_{23}}{-{c}_{12}{c}_{22}{c}_{34}+{c}_{12}{c}_{23}{c}_{33}-{c}_{32}{c}_{13}{c}_{23}}.$  This is a generalization of the example which appears at the end of [Ped, where the intersection of two perpendicular cylinders with a sphere is considered:
$\left\{\begin{array}{ccc}{f}_{1}& =& -1+{t}_{1}^{2}+{t}_{2}^{2}\\ {f}_{2}& =& -1+{t}_{2}^{2}+{t}_{3}^{2}\\ {f}_{3}& =& -1+{t}_{1}^{2}+{t}_{2}^{2}+{t}_{3}^{2},\end{array}$  and the trace of the multiplication by ${t}_{2}^{2}$  is shown to be $8.$

3.2 The denominator of the trace in the torus

In this section we will give an explicit factorization of the denominator of the trace.
The following result is straightforward due to the irreducibility of the resultant of a system of generic polynomials ([GKZ), and the fact that the degree of the numerator of the right hand side of ( 3 ) with respect to the variables ${c}_{ia},{q}_{a}$  is strictly less than the degree of the denominator.
Lemma 3.4. Suppose that ${\mathcal{A}}_{0}={\mathcal{A}}^{\prime }$  (i.e. $\mathcal{A}\subset {\mathcal{A}}^{\prime }$  ). Then either the trace is identically zero, or the right hand side of ( 3 ) is the irreducible representation of the trace as a rational function in $\mathbb{Q}\left({c}_{ia},{p}_{a},{q}_{a}\right).$
Suppose now that $\mathcal{A}$  is not contained in ${\mathcal{A}}^{\prime }.$  As the trace is linear in the monomial expansion of $p,$  it is enough to consider the following situation: $p\left(\mathbf{t}\right)={\mathbf{t}}^{a},$  with $a/\in {\mathcal{A}}^{\prime }.$  Without loss of generality, let $\left\{{\mathcal{A}}_{0},\dots ,{\mathcal{A}}_{j}\right\}$  $\left(j\le k\right)$  be the unique essential subfamily of ${\mathcal{A}}_{0},\dots ,{\mathcal{A}}_{k}$  . Let ${\delta }_{{\mathcal{A}}^{\prime }}:={\mathbf{e}}_{{\mathcal{A}}^{\prime },{\mathcal{A}}_{1},\dots ,{\mathcal{A}}_{j}}\left[L\left({\mathcal{A}}_{0},{\mathcal{A}}_{1},\dots ,{\mathcal{A}}_{j}\right):L\left({\mathcal{A}}^{\prime },{\mathcal{A}}_{1},\dots ,{\mathcal{A}}_{j}\right)\right].$  For each facet of the Minkowski sum ${P}_{1}+\dots +{P}_{j},$  we consider its primitive inward normal vector $\omega$  and define ${\mu }_{\omega }:=min\left\{〈b,\omega 〉,b\in {\mathcal{A}}^{\prime }\right\}-min\left\{〈b,\omega 〉,b\in {\mathcal{A}}_{0}\right\}.$  Observe that ${\mu }_{\omega }\ge 0$  and equality may hold. Set ${\delta }_{\omega }:={\mu }_{\omega }{\mathbf{e}}_{{\mathcal{A}}_{1}^{\omega },\dots ,{\mathcal{A}}_{j}^{\omega }}\left[{\omega }^{\perp }:L\left({\mathcal{A}}_{1}^{\omega },\dots ,{\mathcal{A}}_{j}^{\omega }\right)\right],$  where ${\mathcal{A}}_{i}^{\omega }$  is the set of integer points that lie in the face with inward normal vector $\omega ,$  and ${\omega }^{\perp }$  is the lattice of integer points contained in the hyperplane orthogonal to $\omega$  in $L\left({\mathcal{A}}_{0},{\mathcal{A}}_{1},\dots ,{\mathcal{A}}_{j}\right).$  The main Theorem in [Mingives us the following:
Proposition 3.5. In the situation described above, we have that the denominator of ( 3 ) has the following irreducible factorization:
$Re{s}_{{\mathcal{A}}^{\prime },{\mathcal{A}}_{1},\dots ,{\mathcal{A}}_{k}}\left(q,{f}_{1},\dots ,{f}_{k}{\right)}^{{\delta }_{{\mathcal{A}}^{\prime }}}{\prod }_{\omega }Re{s}_{{\mathcal{A}}_{1}^{\omega },\dots ,{\mathcal{A}}_{k}^{\omega }}\left({f}_{1}^{\omega },\dots ,{f}_{k}^{\omega }{\right)}^{{\delta }_{\omega }}$  where $\omega$  ranges over the primitive inward normal vectors of the facets of ${P}_{1}+\dots +{P}_{j}.$
Example 3.6. Consider the following system
 $\begin{array}{c}\left\{\begin{array}{ccc}{f}_{1}& =& {c}_{11}{t}_{1}+{c}_{12}{t}_{1}{t}_{2}+{c}_{13}{t}_{2}^{2}\\ {f}_{2}& =& {c}_{21}{t}_{2}+{c}_{22}{t}_{1}{t}_{2}+{c}_{23}{t}_{1}^{2}.\end{array}\end{array}$ (5)
We set $q:={q}_{1}+{q}_{2}{t}_{1}+{q}_{3}{t}_{2},a:=\left(2,0\right).$  The Newton polygon ${P}_{1}+{P}_{2}$  is a pentagon whose vertices are $\left(0,3\right),\left(1,3\right),\left(3,1\right),\left(3,0\right),\left(1,1\right).$  The inward normal vectors $\omega$  of this polygon satisfying ${\mu }_{\omega }>0$  are $\left(-1,-1\right)$  and $\left(-1,0\right).$  The facet resultants associated with these edges are ${c}_{22}{c}_{12}-{c}_{13}{c}_{23}$  and ${c}_{23}$  respectively and
 $\begin{array}{c}Re{s}_{{\mathcal{A}}_{0},{\mathcal{A}}_{1},{\mathcal{A}}_{2}}\left(q,{f}_{1},{f}_{2}\right)=Re{s}_{{\mathcal{A}}^{\prime },{\mathcal{A}}_{1},{\mathcal{A}}_{2}}\left(g,{f}_{1},{f}_{2}\right){c}_{23}\left({c}_{22}{c}_{12}-{c}_{13}{c}_{23}\right).\end{array}$ (6)
This is the irreducible decomposition of the denominator of $Trace\left({\text{Times}}_{{\mathbf{t}}^{a}/q}\right).$  Indeed, computing explicitly, we get that its numerator is
 $\begin{array}{c}2{c}_{13}{c}_{11}^{2}{c}_{22}^{3}{q}_{2}{c}_{21}{q}_{1}-3{c}_{13}^{2}{c}_{11}{c}_{21}^{2}{q}_{2}{c}_{23}^{2}{q}_{1}-2{c}_{13}{c}_{11}{c}_{21}{c}_{22}^{2}{q}_{1}^{2}{c}_{23}{c}_{12}+2{c}_{11}^{2}{c}_{21}^{2}{q}_{3}^{2}{c}_{23}{c}_{22}{c}_{12}-{c}_{13}{c}_{11}^{2}{c}_{22}^{4}{q}_{1}^{2}\end{array}$
 $\begin{array}{c}-{c}_{13}{c}_{11}^{2}{c}_{22}^{2}{q}_{2}^{2}{c}_{21}^{2}-{c}_{13}{c}_{11}^{2}{q}_{3}{c}_{23}{q}_{2}{c}_{22}{c}_{21}^{2}-{c}_{13}{c}_{11}^{2}{q}_{3}{c}_{23}{c}_{22}^{2}{q}_{1}{c}_{21}-{c}_{11}^{2}{q}_{3}{c}_{22}^{3}{c}_{12}{q}_{1}{c}_{21}\end{array}$
 $\begin{array}{c}+4{c}_{13}{c}_{11}{q}_{1}{c}_{12}{c}_{23}^{2}{c}_{21}^{2}{q}_{3}+{c}_{13}{c}_{11}{c}_{21}^{3}{q}_{2}^{2}{c}_{23}{c}_{12}-{c}_{13}{q}_{1}^{2}{c}_{12}^{2}{c}_{23}^{2}{c}_{21}^{2}+{c}_{11}^{3}{c}_{22}^{3}{q}_{3}^{2}{c}_{21}+{c}_{13}{c}_{11}{c}_{21}^{2}{q}_{2}{c}_{23}{q}_{1}{c}_{22}{c}_{12}\end{array}$
 $\begin{array}{c}+4{c}_{13}^{2}{c}_{11}{c}_{21}{c}_{22}{q}_{1}^{2}{c}_{23}^{2}-3{c}_{13}{c}_{11}^{2}{c}_{21}^{2}{q}_{3}^{2}{c}_{23}^{2}+{c}_{11}^{2}{q}_{3}{c}_{22}^{2}{c}_{12}{q}_{2}{c}_{21}^{2}-2{c}_{11}{c}_{21}^{2}{q}_{1}{c}_{12}^{2}{c}_{23}{q}_{3}{c}_{22},\end{array}$
which is an irreducible polynomial and does not divide ( 6 ).

4 Traces and Residues in the torus

In this section, we will compare our denominator formulas for traces in the torus with those that can be obtained by applying the results given in [CDS2. As in Section  3 , we will be dealing with a system of $k$  generic Laurent polynomials ${f}_{i}:={\sum }_{a\in {\mathcal{A}}_{i}}{c}_{ia}{\mathbf{t}}^{a}$  , where $i=1,\dots ,k,$  ${\mathcal{A}}_{i}$  is a finite subset of ${\mathbb{Z}}^{k}$  , and the ${c}_{ia}$  's are indeterminates over $\mathbb{Q}.$  For a given Laurent polynomial $p$  in $S:=K\left[{t}_{1},{{t}_{1}}^{-1},\dots ,{t}_{k},{{t}_{k}}^{-1}\right],$  where $K$  was defined in Section  3 , the global residue of the differential form ${\phi }_{p}:=\frac{p}{{f}_{1}\dots {f}_{k}}\frac{d{t}_{1}}{{t}_{1}}\wedge \dots \wedge \frac{d{t}_{k}}{{t}_{k}}$  equals, with our notation,
 $\begin{array}{c}Residu{e}_{f}^{T}\left(p\right):=Trace\left(Time{s}_{\frac{p}{{J}_{f}^{T}}}\right),\end{array}$ (7)
where ${J}_{f}^{T}$  denotes the affine toric Jacobian ${J}_{f}^{T}:=det{\left({t}_{j}\frac{\partial {f}_{i}}{\partial {t}_{j}}\right)}_{1\le i,j\le k}.$  In [CDS2, formulas for the denominator of the rational expression ( 7 ) were proposed. In particular, for any Laurent monomial ${\mathbf{t}}^{a}\in S$  , by replacing $p$  with ${\mathbf{t}}^{a}{J}_{f}^{T}$  in ( 7 ), we get that $Trace\left(Time{s}_{{\mathbf{t}}^{a}}\right)=Residu{e}_{f}^{T}\left({\mathbf{t}}^{a}{J}_{f}^{T}\right),$  and so, the formulas in [CDS2can be used for computing the denominator of the trace of $Time{s}_{{\mathbf{t}}^{a}}.$  Now, we will compare the denominator formula obtained in this way with our results in Section  3 .
Assume that, for $i=1,...,k$  , ${\mathcal{A}}_{i}={\Delta }_{i}\cap {\mathbb{Z}}^{k}$  for an integral polytope ${\Delta }_{i}$  in ${\mathbb{R}}^{k}$  and, without loss of generality, that ${\mathcal{A}}_{i}\subset \left({\mathbb{Z}}_{>0}{\right)}^{k}$  .
Under these assumptions, the support of each of the polynomials ${t}_{j}\frac{\partial {f}_{i}}{\partial {t}_{j}}$  $\left(j=1,...,k\right)$  is ${\mathcal{A}}_{i}$  , and therefore, the support of ${J}_{f}^{T}$  is contained in $\Delta \cap {\mathbb{Z}}^{k}$  , where $\Delta :={\Delta }_{1}+\cdots +{\Delta }_{k}$  .
Thus, in order to obtain a denominator for $Trace\left(Time{s}_{{\mathbf{t}}^{a}}\right)$  it suffices to get denominator formulas for the residues in the torus of those monomials ${\mathbf{t}}^{m}$  with $m\in \left(\Delta \cap {\mathbb{Z}}^{k}\right)+a$  .
First, we introduce some notation. For each facet of $\Delta$  with primitive inward normal vector $\omega$  , let ${a}_{\omega }\in \mathbb{Z}$  be defined as ${a}_{\omega }:=-min\left\{〈b,\omega 〉:b\in \Delta \right\},$  and for every $m\in {\mathbb{Z}}^{k}$  , let
 $\begin{array}{ccc}{\mu }_{\omega }^{-}\left(m\right)& :=& -min\left\{0,〈m,\omega 〉+{a}_{\omega }-1\right\},\end{array}$
 $\begin{array}{ccc}{\delta }_{\omega }^{\prime }\left(m\right)& :=& {\mu }_{\omega }^{-}\left(m\right)\left[{\omega }^{\perp }:L\left({\mathcal{A}}_{1}^{\omega },...,{\mathcal{A}}_{k}^{\omega }\right)\right].\end{array}$
Then, [CDS2,Theorem3.2states that ${\prod }_{\omega }Re{s}_{{\mathcal{A}}_{1}^{\omega },...,{\mathcal{A}}_{k}^{\omega }}\left({f}_{1}^{\omega },...,{f}_{k}^{\omega }{\right)}^{{\delta }_{\omega }^{\prime }\left(m\right)},$  where the product runs over all the primitive inward normal vectors of facets of $\Delta$  , is a denominator for $Residu{e}_{f}^{T}\left({\mathbf{t}}^{m}\right)$  . Therefore, if ${\delta }_{\omega }^{\prime }:=max\left\{{\delta }_{\omega }^{\prime }\left(m\right):m\in \left(\Delta \cap {\mathbb{Z}}^{k}\right)+a\right\},$  the following polynomial is a denominator for $Trace\left(Time{s}_{{\mathbf{t}}^{a}}\right)=Residu{e}_{f}^{T}\left({\mathbf{t}}^{a}{J}_{f}^{T}\right)$  :
${\prod }_{\omega }Re{s}_{{\mathcal{A}}_{1}^{\omega },...,{\mathcal{A}}_{k}^{\omega }}\left({f}_{1}^{\omega },...,{f}_{k}^{\omega }{\right)}^{{\delta }_{\omega }^{\prime }}.$  Let us estimate the exponents ${\delta }_{\omega }^{\prime }$  . For $m\in \left(\Delta \cap {\mathbb{Z}}^{k}\right)+a$  , write $m={m}_{\Delta }+a$  with ${m}_{\Delta }\in \Delta \cap {\mathbb{Z}}^{k}$  . We have that $〈m,\omega 〉+{a}_{\omega }-1=〈{m}_{\Delta },\omega 〉+{a}_{\omega }+〈a,\omega 〉-1\ge 〈a,\omega 〉-1$  .
We will consider two cases separetely:
• $〈a,\omega 〉>0$  : for every $m\in \left(\Delta \cap {\mathbb{Z}}^{k}\right)+a$  , the previous inequality implies that $〈m,\omega 〉+{a}_{\omega }-1\ge 0$  , and so, ${\mu }_{\omega }^{-}\left(m\right)=0$  . Therefore, ${\delta }_{\omega }^{\prime }=0$  .
• $〈a,\omega 〉\le 0$  : taking ${m}_{\Delta }\in \Delta \cap {\mathbb{Z}}^{k}$  so that $〈{m}_{\Delta },\omega 〉+{a}_{\omega }=0$  , for $m:={m}_{\Delta }+a$  , we get $〈m,\omega 〉+{a}_{\omega }-1=〈a,\omega 〉-1\le -1$  . Then, ${\mu }_{\omega }^{-}\left(m\right)=1-〈a,\omega 〉\ge 1$  and therefore, ${\delta }_{\omega }^{\prime }=\left(1-〈a,\omega 〉\right)L\left({\mathcal{A}}_{1}^{\omega },...,{\mathcal{A}}_{k}^{\omega }\right)$  .
We deduce the following formula for a denominator of $Trace\left(Time{s}_{{\mathbf{t}}^{a}}\right)$  :
 $\begin{array}{c}{\prod }_{\omega :〈a,\omega 〉\le 0}Re{s}_{{\mathcal{A}}_{1}^{\omega },...,{\mathcal{A}}_{k}^{\omega }}\left({f}_{1}^{\omega },...,{f}_{k}^{\omega }{\right)}^{\left(1-〈a,\omega 〉\right)L\left({\mathcal{A}}_{1}^{\omega },...,{\mathcal{A}}_{k}^{\omega }\right)}.\end{array}$ (8)
Finally, we will restate the result in Proposition  3.5 in this context: here $\mathcal{A}=\left\{a\right\}$  , ${\mathcal{A}}^{\prime }=\left\{0\right\}$  and ${\mathcal{A}}_{0}=\left\{0,a\right\}$  . The first resultant appearing in the factorization equals $1$  . On the other hand, it follows from the definition that ${\mu }_{\omega }=-min\left\{0,〈a,\omega 〉\right\}$  , and so, ${\delta }_{\omega }=-〈a,\omega 〉L\left({\mathcal{A}}_{1}^{\omega },...,{\mathcal{A}}_{k}^{\omega }\right)$  if $〈a,\omega 〉\le 0$  and ${\delta }_{\omega }=0$  otherwise. We conclude that
 $\begin{array}{c}{\prod }_{\omega :〈a,\omega 〉\le 0}Re{s}_{{\mathcal{A}}_{1}^{\omega },...,{\mathcal{A}}_{k}^{\omega }}\left({f}_{1}^{\omega },...,{f}_{k}^{\omega }{\right)}^{-〈a,\omega 〉L\left({\mathcal{A}}_{1}^{\omega },...,{\mathcal{A}}_{k}^{\omega }\right)}\end{array}$ (9)
is a denominator for $Trace\left(Time{s}_{{\mathbf{t}}^{a}}\right)$  .
By comparing ( 9 ) with ( 8 ) we get a slightly improvement on the exponent of each of the factors in the denominator.

5 Multidimensional Residues in ${\mathbb{C}}^{n}$

In the case where the underlying variety is the zero locus of a regular sequence of $n$  polynomials in $\mathbb{C}\left[{x}_{1},\dots ,{x}_{n}\right]$  without zeroes in the infinity, results of Jouanolou on generalized discriminants ([Jou) will allow us to use our formulas in order to recover known results about residues ([CDS1, CDS2).

5.1 The denominator of the residue

Let $f:=\left({f}_{1},...,{f}_{n}\right)$  with ${f}_{i}:={\sum }_{|a|\le {d}_{i}}{c}_{ia}{x}^{a}$  , $i=1,...,n$  , be a generic system of polynomials in $\mathbb{K}\left[{x}_{1},...,{x}_{n}\right]$  of respective degrees ${d}_{1},...,{d}_{n}$  . Set ${J}_{f}:=det\left(\partial {f}_{i}/\partial {x}_{j}\right)$  for the Jacobian determinant of the system. Let us observe that $deg\left({J}_{f}\right)=\rho :={\sum }_{i=1}^{n}{d}_{i}-n$  .
For every $\beta \in \left({\mathbb{Z}}_{\ge 0}{\right)}^{n}$  , the global residue associated with the data $\left({x}^{\beta },f\right)$  can be obtained as $Residu{e}_{f}\left({x}^{\beta }\right)=Trace\left(Time{s}_{{x}^{\beta }/{J}_{f}}\right),$  where the linear maps in the right hand side of the equation are defined in the quotient ring $\mathbb{K}\left[{x}_{1},...,{x}_{n}\right]/〈{f}_{1},...,{f}_{n}〉$  . Applying Theorem  3 we obtain the following expression for the residue as a rational function in the coefficients of $f$  :
$Residu{e}_{f}\left({x}^{\beta }\right)=\frac{\frac{\partial Re{s}_{D,{d}_{1},...,{d}_{n}}}{\partial {c}_{0\beta }}\left({J}_{f},{f}_{1},\dots ,{f}_{n}\right)}{Re{s}_{D,{d}_{1},...,{d}_{n}}\left({J}_{f},{f}_{1},\dots ,{f}_{n}\right)},$  where $D:=max\left\{|\beta |,\rho \right\}$  , $Re{s}_{D,{d}_{1},...,{d}_{n}}$  is the classic resultant of a generic system of $n+1$  polynomials with respective degrees $D,{d}_{1},...,{d}_{n}$  , and $\frac{\partial }{\partial {c}_{0\beta }}$  stands for the derivative with respect to the coefficient of the monomial ${x}^{\beta }$  in the first polynomial.
Due to Proposition  3.5 , if ${f}_{i}^{0}:={\sum }_{|a|={d}_{i}}{c}_{ia}{x}^{a}$  is the homogeneous part of degree ${d}_{i}$  of the polynomial ${f}_{i}$  , $i=1,...,n$  , the denominator of this expression factors as $Re{s}_{D,{d}_{1},...,{d}_{n}}\left({J}_{f},{f}_{1},...,{f}_{n}\right)=Re{s}_{{d}_{1},...,{d}_{n}}\left({f}_{1}^{0},...,{f}_{n}^{0}{\right)}^{D-\rho }Re{s}_{\rho ,{d}_{1},...,{d}_{n}}\left({J}_{f},{f}_{1},...,{f}_{n}\right).$  Now, according to [Jou, the following identity holds:
$Re{s}_{\rho ,{d}_{1},...,{d}_{n}}\left({J}_{f},{f}_{1},...,{f}_{n}\right)=Re{s}_{{d}_{1},...,{d}_{n}}\left({f}_{1}^{0},...,{f}_{n}^{0}\right)Disc\left(f\right).$  Here, $Disc\left(f\right)$  denotes the discriminant of the polynomial system $f=\left({f}_{1},...,{f}_{n}\right)$  . Then, we can factor the denominator of the residue further.
Let us recall that $Disc\left(f\right)$  is an irreducible polynomial in the coefficients of $f$  which vanishes if and only if the system $f$  has a multiple root or has infinitely many roots, and that the same happens to the polynomial appearing in the numerator of the residue (this is due to the so called “Principle of Continuity”, see [GH,p. 657). Being $Disc\left(f\right)$  an irreducible polynomial, we deduce that
 $\begin{array}{c}Residu{e}_{f}\left({x}^{\beta }\right)=\frac{{A}_{\beta }\left(f\right)}{Re{s}_{{d}_{1},...,{d}_{n}}\left({f}_{1}^{0},...,{f}_{n}^{0}{\right)}^{D-\rho +1}},\end{array}$ (10)
where ${A}_{\beta }\left(f\right)\in \mathbb{K}\left[{c}_{ia}\right]$  is a polynomial in the coefficients of ${f}_{1},...,{f}_{n}$  .

5.2 An algebraic proof of the Euler-Jacobi vanishing theorem

The following result is very well known in the literature, see for instance [Macor [AGV,Chapter 5 ,Corollary 4 . We will recover it by using Identity ( 10 ) and algebraic methods.
Theorem 5.1 (Euler-Jacobi Formula). For a system $f=\left({f}_{1},\dots ,{f}_{n}\right)$  as before, and any polynomial $h$  of degree less than the degree of the Jacobian of the system, $Residu{e}_{f}\left(h\right)=0.$
• Proof. It suffices to show that $Residu{e}_{f}\left({x}^{\beta }\right)=0$  for every monomial of degree at most $\rho :=deg\left({J}_{f}\right)$  .
For $i=1,...,n$  , let ${f}_{t,i}:={\sum }_{|a|\le {d}_{i}}{c}_{ia}{t}^{{d}_{i}-|a|}{x}^{a}\in \mathbb{K}\left[t\right]\left[{x}_{1},...,{x}_{n}\right]$  and let ${f}_{t}$  be the generic system ${f}_{t}:=\left({f}_{t,1},...,{f}_{t,n}\right)$  .
Fix $\beta \in \left({\mathbb{Z}}_{\ge 0}{\right)}^{n}$  with $|\beta |<\rho$  . In what follows, we will relate the global residue of ${x}^{\beta }$  with respect to the original system $f$  with its global residue with respect to the system ${f}_{t}$  .
First, let us observe that $\left\{\xi =\left({\xi }_{1},...,{\xi }_{n}\right)\in {\mathbb{A}}^{n}:{f}_{t}\left(\xi \right)=0\right\}=\left\{t\eta =\left(t{\eta }_{1},...,t{\eta }_{n}\right)\in {\mathbb{A}}^{n}:f\left(\eta \right)=0\right\}$  and that ${J}_{{f}_{t}}\left(t\eta \right)={t}^{\rho }{J}_{f}\left(\eta \right)$  , which implies that $Residu{e}_{{f}_{t}}\left({x}^{\beta }\right)={\sum }_{{f}_{t}\left(\xi \right)=0}\frac{{\xi }^{\beta }}{{J}_{{f}_{t}}\left(\xi \right)}={\sum }_{f\left(\eta \right)=0}\frac{\left(t\eta {\right)}^{\beta }}{{t}^{\rho }{J}_{f}\left(\eta \right)}={t}^{|\beta |-\rho }{\sum }_{f\left(\eta \right)=0}\frac{{\eta }^{\beta }}{{J}_{f}\left(\eta \right)}={t}^{|\beta |-\rho }Residu{e}_{f}\left({x}^{\beta }\right).$  On the other hand, taking into account that ${f}_{i}^{0}={f}_{t,i}^{0}$  , Identity ( 10 ) applied to system ${f}_{t}$  states that $Residu{e}_{{f}_{t}}\left({x}^{\beta }\right)=\frac{{A}_{\beta }\left({f}_{t}\right)}{Re{s}_{{d}_{1},...,{d}_{n}}\left({f}_{1}^{0},...,{f}_{n}^{0}{\right)}^{D-\rho +1}}.$  Finally, combining the previous identities we deduce:
${A}_{\beta }\left({f}_{t}\right)={t}^{|\beta |-\rho }{A}_{\beta }\left(f\right).$  Now, ${A}_{\beta }\left({f}_{t}\right)$  is a polynomial in $\mathbb{K}\left[t,{c}_{ia}\right]$  , while ${A}_{\beta }\left(f\right)\in \mathbb{K}\left[{c}_{ia}\right]$  does not depend on $t$  and $|\beta |-\rho <0$  . Then, the above equality is only possible if ${A}_{\beta }\left(f\right)=0$  and so, $Residu{e}_{f}\left({x}^{\beta }\right)=0$  .
References

1. Alonso, M.-E.; Becker, E.; Roy, M.-F.; Wörmann, T. Zeros, multiplicities, and idempotents for zero-dimensional systems. Algorithms in algebraic geometry and applications (Santander, 1994), 1–15, Progr. Math., 143, Birkhäuser, Basel, 1996.
2. Arnold, V. I.; Guseĭn-Zade, S. M.; Varchenko, A. N. Singularities of differentiable maps. Vol. I. The classification of critical points, caustics and wave fronts. Translated from the Russian by Ian Porteous and Mark Reynolds. Monographs in Mathematics, 82. Birkhäuser Boston, Inc., Boston, MA, 1985.
3. Aĭzenberg, I. A.; Yuzhakov, A. P. Integral representations and residues in multidimensional complex analysis. Translated from the Russian by H. H. McFaden. Translation edited by Lev J. Leifman. Translations of Mathematical Monographs, 58. American Mathematical Society, Providence, RI, 1983.
4. Becker, E.; Wöermann, T. On the trace formula for quadratic forms. Recent advances in real algebraic geometry and quadratic forms (Berkeley, CA, 1990/1991; San Francisco, CA, 1991), 271–291, Contemp. Math., 155, Amer. Math. Soc., Providence, RI, 1994.
5. Cardinal, J. P.; Mourrain, B. Algebraic approach of residues and applications. The mathematics of numerical analysis (Park City, UT, 1995), 189–210, Lectures in Appl. Math., 32, Amer. Math. Soc., Providence, RI, 1996
6. Cattani, E.; Dickenstein, A.; Sturmfels, B. Computing multidimensional residues. Algorithms in algebraic geometry and applications (Santander, 1994), 135–164, Progr. Math., 143, Birkhäuser, Basel, 1996.
7. Cattani, E.; Dickenstein, A.; Sturmfels, B. Residues and resultants. J. Math. Sci. Univ. Tokyo 5 (1998), No. 1, 119–148.
8. Cox, D.; Little, J.; O'Shea, D. Using algebraic geometry. Graduate Texts in Mathematics, 185. Springer-Verlag, New York, 1998.
9. Díaz-Toca, Gema M.; González-Vega, L. An explicit description for the triangular decomposition of a zero-dimensional ideal through trace computations. Green, Edward L. (ed.) et al., Symbolic computation: Solving equations in algebra, geometry, and engineering. Proceedings of an AMS-IMS-SIAM joint summer research conference, Mount Holyoke College, South Hadley, MA, USA, June 11-15, 2000. Providence, RI: AMS, American Mathematical Society. Contemp. Math. 286, 21-35 (2001).
10. Fitchas, N.; Giusti, M.; Smietanski, F. Sur la complexité du théorème des zéros. Approximation and optimization in the Caribbean, II (Havana, 1993), 274–329, Lang, Frankfurt am Main, 1995.
11. Gel'fand, I. M.; Kapranov, M. M.; Zelevinsky, A. V. Discriminants, resultants, and multidimensional determinants. Mathematics: Theory & Applications. Birkhäuser Boston, Inc., Boston, MA, 1994. x+523 pp.
12. Griffiths, Phillip; Harris, Joseph. Principles of algebraic geometry. Pure and Applied Mathematics. Wiley-Interscience [John Wiley & Sons], New York, 1978.
13. Jeronimo, G.; Krick, T.; Sabia, J.; Sombra, M. The computational complexity of the Chow form. Found. Comput. Math. 4, No. 1, 41–117 (2004).
14. Jouanolou, J.P. Course DEA, University of Strasbourg.
15. Kunz, E. Kähler differentials. Advanced Lectures in Mathematics. Friedr. Vieweg & Sohn, Braunschweig, 1986.
16. F. Macaulay. The algebraic theory of modular systems. Cambridge University Press, 1916.
17. Minimair, M. Sparse resultant under vanishing coefficients. J. Algebraic Combin. 18 (2003), no. 1, 53–73.
18. Pedersen, Paul. Calculating multidimensional symmetric functions using Jacobi's formula. Applied algebra, algebraic algorithms and error-correcting codes, Proc. 9th Int. Symp., AAECC-9, New Orleans/LA (USA) 1991, Lect. Notes Comput. Sci. 539, 304-317 (1991).
19. Pedersen, P.; Sturmfels, B. Product formulas for resultants and Chow forms. Math. Z. 214 (1993), No. 3, 377–396.
20. Philippon, P. Critères pour l'indépendance algébrique. Inst. Hautes Études Sci. Publ. Math. No. 64 (1986), 5–52.
21. Sabia, J.; Solernó, P. Bounds for traces in complete intersections and degrees in the Nullstellensatz. Appl. Algebra Engrg. Comm. Comput. 6 (1995), no. 6, 353–376.
22. Sturmfels, B. On the Newton polytope of the resultant. J. Algebr. Comb. 3, No. 2, 207–236 (1994).
23. Tsikh, A. K. Multidimensional residues and their applications. Translated from the 1988 Russian original by E. J. F. Primrose. Translations of Mathematical Monographs, 103. American Mathematical Society, Providence, RI, 1992.

Carlos D'Andrea: Department of Mathematics, University of California, Berkeley, CA 94720, USA. E-mail: cdandrea@math.berkeley.edu.
Gabriela Jeronimo: Departamento de Matemática, FCEyN, Universidad de Buenos Aires, Ciudad Universitaria, 1428 Buenos Aires, Argentina. E-mail: jeronimo@dm.uba.ar.