Pith. sign in

REVIEW 2 major objections 5 minor 39 references

Support bound for differential elimination in polynomial dynamical systems

T0 review · 2 major / 5 minor · reviewed 2026-08-07 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read The paper proves an explicit degree-only support bound for the minimal differential equation obtained by eliminating hidden state from polynomial dynamical systems, which makes differential elimination feasible by evaluation-interpolation.

desk verdict A genuine extension of the support bound to arbitrary polynomial outputs and symbolic parameters, with a solid proof that has a few typos and one presentation gap worth fixing before publication. read the letter →

arxiv 2506.08824 v2 pith:KUMF4IO4 submitted 2025-06-10 cs.SC math.AGmath.CA

Yulia Mukhina , Gleb Pogudin This is my paper
classification cs.SCmath.AGmath.CA MSC 12H0568W3013P10
keywords differentialeliminationNewtonpolytopesupportboundevaluation-interpolationpolynomialdynamicalsystemsminimalequationparametric
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

This paper proves that eliminating the hidden state from a polynomial parametric dynamical system $x' = g(\mu, x)$ with a polynomial observation $y = f(\mu,x)$ has a finite, a priori computable shape. For every monomial $\mu^{\ell} y^{e_0}(y')^{e_1}\cdots(y^{(\nu)})^{e_{\nu}}$ of the minimal polynomial $f_{\min}$, the exponents satisfy three explicit inequalities, (4)-(6), whose right-hand sides depend only on the degrees of $f$ and $g$ and on an order bound $\nu$ that can always be taken to be the dimension $n$. The support of $f_{\min}$ is therefore contained in an explicitly bounded Newton polytope---the convex hull of the possible exponent vectors---that can be computed before any elimination. This turns the problem into an evaluation-interpolation computation: one knows which monomials can appear, samples truncated power series solutions, and solves a linear system for the coefficients, avoiding the intermediate expression swell that limits classical algorithms. The result covers arbitrary polynomial observation functions and parameters, and the experiments show the bound is often close to sharp and lets the implementation finish dense parametric instances that other software cannot handle.

What carries the argument

The load-bearing objects are the minimal polynomial $f_{\min}$ and its Newton polytope in the variables $\mu, y, y',\dots,y^{(\nu)}$. Three ingredients work together: the Lie-derivative flattening $y^{(i)}=L_g^i(f)$, which turns differential constraints into polynomial equations; a generic univariate substitution $y^{(i)}\mapsto p_i(z_i)$, chosen through a lemma from the earlier paper so that the substituted minimal polynomial is square-free (no repeated factors); and the multihomogeneous Bezout bound, applied after bihomogenizing the resulting system, to bound the degree of that substituted polynomial. Lemmas 3 and 4 arrange that the relevant fibers are finite and nonempty, and Lemma 5 converts the Bezout count into the explicit weighted-exponent inequalities (4)-(6).

What would settle it

Compute the exact minimal polynomial for a small dense parametric system, say $n=2$, $r=1$, $d_x=D_x=d_\mu=D_\mu=2$, with an independent elimination routine, and test every monomial against inequalities (4)-(6); a single monomial outside the predicted polytope would refute Theorem 1. A more targeted test would search directly for a system in which $\varphi(f_{\min})$ is never square-free for any generic choice of the $p_i$ with the prescribed degrees.

Watch

Extended reading notes

Core claim

The central claim is Theorem 1: given polynomials $g_1,\dots,g_n,f$ in $K[\mu,x]$ with $d_\alpha = \deg_\alpha f$ and $D_\alpha = \max_i \deg_\alpha g_i$ for $\alpha = \mu$ or $\alpha = x$, and any positive integer $\nu$ with $\operatorname{ord} f_{\min} \le \nu$ (one may always take $\nu = n$), every monomial $\mu^{\ell} y^{e_0}(y')^{e_1}\cdots(y^{(\nu)})^{e_{\nu}}$ of the minimal polynomial $f_{\min}$ satisfies inequalities (4), (5), and (6). Inequality (6) implies the non-parametric Corollary 1, bounding $\sum_{k=0}^{\nu} (d+k(D-1))e_k$ by $\prod_{k=0}^{\nu}(d+k(D-1))$. The proof reduces the differential problem to polynomial elimination using the Lie-derivative identities $y^{(i)} = L_g^i(f)$, then bounds the degree of a generically substituted $f_{\min}$ by the multihomogeneous Bezout bound and dimension arguments. The consequence is that the support of the minimal differential equation is finite and can be computed before performing any elimination, which is exactly what the evaluation-interpolation algorithm needs.

Load-bearing premise

The proof rests on an unproved lemma from the earlier paper: for the prescribed degrees, generic univariate substitutions of the derivatives make the substituted minimal polynomial square-free, meaning it has no repeated factors; if that simultaneous choice fails, the degree estimate and therefore the support bound do not follow.

Editorial extensions

If this is right

  • Before any elimination, the finite support of $f_{\min}$ can be listed from the degrees of $f$ and $g$, so the evaluation-interpolation algorithm terminates and is correct for every input with known degrees.
  • Arbitrary polynomial observation functions are covered, so nonlinear quantities such as products of coordinates---for example the product of $\tan$ and $\tanh$, each satisfying a polynomial ODE---can be eliminated directly, not just single coordinates.
  • Parameters $\mu$ remain symbolic in the bound, giving a principled handle on elimination problems with unknown constant parameters, the setting needed for structural identifiability.
  • The numerical experiments show the bound is sharp in the non-parametric dense cases tested and that in parametric cases its accuracy increases as the degrees grow, so the support bound is a reliable predictor of the size of $f_{\min}$.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Editorial: the same Lie-derivative and Bezout mechanism should extend to rational observation functions after clearing denominators, with adjusted degree parameters, because the flattening identities control degrees exactly.
  • Editorial: the near-100% density observed in the experiments hints that for generic dense inputs of fixed degrees the minimal polynomial fills its Newton polytope entirely; if proved, the bound would become an exact predictor of elimination output size rather than merely an upper bound.
  • Editorial: because the bound tracks monomials in the parameters, it could be used to precompute which parameter monomials can appear in an input-output equation, a step that structural identifiability pipelines currently discover only after expensive elimination.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

2 major / 5 minor

Summary. This paper studies the differential elimination problem for polynomial parametric dynamical systems x'=g(μ,x) with a polynomial observation y=f(μ,x). The main result, Theorem 1, gives three explicit inequalities (4)–(6) bounding the support (Newton polytope) of the minimal differential polynomial fmin satisfied by y, with bounds depending only on the degrees of f and g and the order bound ν. The proof reduces the differential problem to polynomial elimination by substituting y^(i)=p_i(z_i) for generic univariate polynomials p_i, applies dimension theory and multihomogeneous Bézout bounds, and uses two lemmas from the companion paper [29]. The authors validate the bound experimentally and present an implementation, DiffMinPoly, that outperforms existing differential elimination software on several benchmark systems.

Significance. The result is significant because it provides, for the first time, a pre-computational bound on the support of the minimal differential equation for arbitrary polynomial observations with symbolic parameters. This enables evaluation-interpolation algorithms that avoid intermediate expression swell, as demonstrated by the experiments. The paper includes reproducible code and careful experimental reporting. If the proof issues are resolved, the contribution is a solid step toward applying differential elimination to structural identifiability and D-algebraic functions. The main caveats are the proof's dependence on unstated lemmas from a companion preprint and several dimension typos that obscure the argument.

major comments (2)
  1. [Section 7, Case 1 and Section 5, Corollary 2] The proof contains dimension errors that make the argument incoherent as printed. In Case 1, the ideal \tilde I is generated by ν+1 polynomials in the ring K[μ,x,z] with z=(z_0,...,z_ν), so for a regular sequence the dimension is dim(\tilde I)=r+n, not 2n+r−ν. The projection onto (μ,z) has target A^{r+ν+1}, not A^{n+r+1}, and Y=V(\tilde fmin) is a hypersurface of dimension r+ν, not n+r. The subsequent application of Lemma 3 with an affine space L⊂A^n of codimension n−ν requires D−d = n−ν, i.e., dim Y = r+ν, so the printed 'dim Y = n+r' is inconsistent with the claimed slicing. In Corollary 2, the map φ is written as A^n→A^{s+1} with dim(im φ)=s, but since f and L_g(f) depend on the parameters μ, the map should be A^{r+n}→A^{r+s+1} and its image dimension is r+s. These are fixable typos, but they must be corrected before the proof of Theorem 1 can be followed.
  2. [Section 7] The proof relies on two lemmas from the companion paper [29] that are neither stated nor proved here. At the start of Section 7, [29, Lemma 4] is invoked to assert that generic univariate polynomials p_i make \tilde fmin := φ(fmin) square-free; this square-freeness is needed to apply Lemma 5 of the present paper in Case 1. In Case 2, [29, Lemma 5] supplies the product bound deg \tilde fmin ≤ ∏_{i=0}^ν deg \tilde f_i, which is exactly the right-hand side of inequality (5). Since [29] treated only the case y=x_1 without symbolic parameters, the manuscript should either state these lemmas and verify that their hypotheses remain valid for arbitrary polynomial f and parameters μ, or prove them in an appendix. As written, the proof of Theorem 1 is not self-contained and its validity depends on assertions in an unpublished companion paper.
minor comments (5)
  1. [Section 3, Theorem 1] In inequalities (4) and (5), the sum over the μ-exponents is written as ∑_{i=0}^r ℓ_i, but ℓ_i is defined for i=1,...,r; it should be ∑_{i=1}^r ℓ_i.
  2. [Section 7, Case 1] The sentence 'Let ω_i = dµ + iDµ for i = 1, . . . , ν' should read 'for i = 0, . . . , ν', since ω_0 is used in the subsequent degree computation.
  3. [Table 3(a)] In the row [Dx,dx]=[1,1] with |μ|=1, the columns 'NP of fmin' and 'fmin' report 10 and 20, respectively, which is inconsistent with the statement that in all experiments the minimal polynomial is dense in its Newton polytope (a polynomial cannot have more monomials than lattice points in its Newton polytope). This appears to be a typo that should be corrected.
  4. [Section 7, Case 1] The expression 'deg_{μ,y} \tilde f_k' should presumably be 'deg_{μ,z} \tilde f_k', since the y^{(i)} variables have been replaced by z_i.
  5. [General proof] The theorem statement uses ν both as an arbitrary upper bound on ord fmin and, in the proof, as the exact order. While the inequalities for a larger ν follow from those at the exact order, the overloaded notation is confusing and should be clarified.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the support bound is a derived mathematical inequality, not a fitted or self-referential prediction.

full rationale

The paper's central result, Theorem 1, is a support bound for the minimal differential polynomial fmin. It is proved by degree estimates on the elimination ideal after a substitution y^(i) -> p_i(z_i). The proof reuses two technical lemmas from the authors' prior work [29] (a generic square-freeness statement and a degree-product bound), but those are ordinary citations of established mathematical results, not definitions of the target quantities. No parameter appearing in the bound is fitted to the data being bounded, and the experimental section independently computes fmin for random models and compares it to the formula. The algorithm in Section 8 uses the bound as an ansatz support and then solves linear systems for coefficients; that is a computational method, not a prediction that has been fitted into existence. The reliance on [29, Lemmas 4 and 5], and on the order bound from [20], is a verification/dependency concern, not circularity, because the present claim is not equivalent to any of these inputs by construction. The derivation is self-contained as a proof argument, with only standard external lemmas invoked.

Assumptions & free parameters 0 free parameters · 5 assumptions · 0 invented entities

No parameters are fitted to data: the degrees dμ, Dμ, dx, Dx and the counts n, r are inputs, and the evaluation-interpolation coefficients are solved from a determined linear system rather than tuned. The paper introduces no new mathematical entities beyond standard objects such as fmin, Lie derivatives, and Newton polytopes. The main load is prior lemmas from the same group and standard algebraic geometry, which are legitimate reuse rather than invented structure.

assumptions (5)
  • standard math K is a differential field of characteristic zero, with the usual Leibniz derivation properties.
    All differential-algebraic constructions in the paper, including Definition 1 and Notation 3, assume this setting.
  • domain assumption The ideal I_{g,f} = (x'-g, y-f)^{(∞)} is prime for polynomial g,f.
    Primeness is needed so the elimination ideal has the unique minimal polynomial fmin used throughout; cited to [27] and [20, Lemma 3.2].
  • domain assumption The order of fmin is at most the state dimension n, so ν = n can always be used in Theorem 1.
    Invoked in Section 7 via [20, Theorem 3.16 and Corollary 3.21]; also follows from transcendence degree considerations.
  • standard math Generic slicing, finite-fiber, and multihomogeneous Bezout facts from Section 6 hold as stated.
    Lemma 3, Lemma 4, and Lemma 5 are the geometric engine of the proof; Lemma 5 quotes the Bezout bound from [28].
  • domain assumption The substitution lemma [29, Lemma 4] makes φ(fmin) square-free for generic p_i of prescribed degrees, and [29, Lemma 5] bounds deg φ(fmin) by a product of degrees.
    These lemmas are imported from the authors' earlier paper and are not reproduced; they are load-bearing in the proof of Theorem 1.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Support bound for differential elimination in polynomial dynamical systems." pith.science (2026). https://pith.science/paper/KUMF4IO4

@misc{pith2026250608824,
  author       = {Pith},
  title        = {Pith review of: Support bound for differential elimination in polynomial dynamical systems},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/KUMF4IO4}},
  note         = {Machine review of arXiv:2506.08824}
}
abstract

We study an important special case of the differential elimination problem: given a polynomial parametric dynamical system $\mathbf{x}' = \mathbf{g}(\boldsymbol{\mu}, \mathbf{x})$ and a polynomial observation function $y = f(\boldsymbol{\mu}, \mathbf{x})$, find the minimal differential equation satisfied by $y$. In our previous work, for the case $y = x_1$, we established a bound on the support of such a differential equation for the non-parametric case and shown that it can be turned into an algorithm via the evaluation-interpolation approach. The main contribution of the present paper is a generalization of the aforementioned result in two directions: to allow any polynomial function $y = f(\mathbf{x})$, not just a single coordinate, and to allow $\mathbf{g}$ and $f$ depend on unknown symbolic parameters. We conduct computation experiments to evaluate the accuracy of our new bound and show that the approach allows to perform elimination for some cases out of reach for the state of the art software.

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

39 extracted references · 22 canonical work pages

  1. [29]

    Mukhina, Y., Pogudin, G.: Projecting dynamical systems via a support bound (2025), https://arxiv.org/abs/2501.13680

    work page Pith review arXiv 2025

  2. [20]

    Communications on Pure and Applied Mathematics73(9), 1831–1879 (2020), https://doi.org/10.1002/cpa.21921

    Hong, H., Ovchinnikov, A., Pogudin, G., Yap, C.: Global identifiability of differen- tial models. Communications on Pure and Applied Mathematics73(9), 1831–1879 (2020), https://doi.org/10.1002/cpa.21921

    doi:10.1002/cpa.21921 2020

  3. [1]

    Journal of Symbolic Computation128, 102377 (May 2025), http://dx.doi

    Ait El Manssour, R., Sattelberger, A.L., Teguia Tabuguia, B.: D-algebraic func- tions. Journal of Symbolic Computation128, 102377 (May 2025), http://dx.doi. org/10.1016/j.jsc.2024.102377

    arXiv 2025

  4. [2]

    In: Computer Algebra in Scientific Comput- ing, pp

    Bächler, T., Gerdt, V., Lange-Hegermann, M., Robertz, D.: Thomas decomposition of algebraic and differential systems. In: Computer Algebra in Scientific Comput- ing, pp. 31–54 (2010), https://doi.org/10.1007/978-3-642-15274-0_4

    doi:10.1007/978-3-642-15274-0_4 2010

  5. [3]

    Journal of Symbolic Compu- tation 47(10), 1233–1266 (2012), https://doi.org/10.1016/j.jsc.2011.12.043

    Bächler, T., Gerdt, V., Lange-Hegermann, M., Robertz, D.: Algorithmic Thomas decomposition of algebraic and differential systems. Journal of Symbolic Compu- tation 47(10), 1233–1266 (2012), https://doi.org/10.1016/j.jsc.2011.12.043

    doi:10.1016/j.jsc.2011.12.043 2012

  6. [4]

    In: Proceedings of the 1995 International Symposium on Symbolic and Algebraic Computation - ISSAC’95 (1995), https: //doi.org/10.1145/220346.220367

    Boulier, F., Lazard, D., Ollivier, F., Petitot, M.: Representation for the radical of a finitely generated differential ideal. In: Proceedings of the 1995 International Symposium on Symbolic and Algebraic Computation - ISSAC’95 (1995), https: //doi.org/10.1145/220346.220367

    work page arXiv 1995

  7. [5]

    univ-lille.fr/francois-boulier/logiciels/blad/

    Boulier, F.: BLAD: Bibliothèques Lilloises d’Algèbre Différentielle, https://pro. univ-lille.fr/francois-boulier/logiciels/blad/

    work page

  8. [6]

    Boulier, F.: DifferentialAlgebra, https://codeberg.org/francois.boulier/ DifferentialAlgebra/

    work page

Show all 39 references
  1. [7]

    In: Gröbner Bases in Symbolic Analysis, p

    Boulier, F.: Differential elimination and biological modelling. In: Gröbner Bases in Symbolic Analysis, p. 109–138. De Gruyter (2007), http://dx.doi.org/10.1515/ 9783110922752.109 Support bound for differential elimination in polynomial dynamical systems 19

    2007

  2. [8]

    Applicable Algebra in Engineering, Communication and Computing 20(1), 73–121 (2009), https://doi.org/10.1007/ s00200-009-0091-7

    Boulier, F., Lazard, D., Ollivier, F., Petitot, M.: Computing representations for radicals of finitely generated differential ideals. Applicable Algebra in Engineering, Communication and Computing 20(1), 73–121 (2009), https://doi.org/10.1007/ s00200-009-0091-7

    2009

  3. [9]

    Hermann (1950)

    Bourbaki, N.: Éléments de mathématique: Les structures fondamentales de l’analyse; Algèbre; Polynomes et fractions rationnelles; Chapitre 5: Corps com- mutatifs. Hermann (1950)

    1950

  4. [10]

    In: International Symposium on Symbolic and Algebraic Manipulation

    Buchberger, B.: A criterion for detecting unnecessary reductions in the construc- tion of gröbner-bases. In: International Symposium on Symbolic and Algebraic Manipulation. pp. 3–21. Springer (1979)

    1979

  5. [11]

    Cox, D., Little, J., O’Shea, D., Sweedler, M.: Ideals, varieties, and algorithms, vol. 3. Springer (1997)

    1997

  6. [12]

    SIAM Journal on Applied Algebra and Geometry7(1), 194–235 (Mar 2023), http: //dx.doi.org/10.1137/22M1469067

    Dong, R., Goodbrake, C., Harrington, H.A., Pogudin, G.: Differential elimination for dynamical models via projections with applications to structural identifiability. SIAM Journal on Applied Algebra and Geometry7(1), 194–235 (Mar 2023), http: //dx.doi.org/10.1137/22M1469067

    2023 doi

  7. [13]

    Eisenbud, D.: Commutative Algebra, Graduate Texts in Mathematics, vol. 150. Springer, New York, NY (1995). https://doi.org/10.1007/978-1-4612-5350-1

    1995 doi

  8. [14]

    In: Proceedings of the 2017 ACM International Symposium on Symbolic and Algebraic Computa- tion

    Fieker, C., Hart, W., Hofmann, T., Johansson, F.: Nemo/Hecke: Computer algebra and number theory packages for the Julia programming language. In: Proceedings of the 2017 ACM International Symposium on Symbolic and Algebraic Computa- tion. p. 157–164. ISSAC ’17 (2017), https://...

    2017

  9. [15]

    In: Polytopes—combinatorics and computation

    Gawrilow, E., Joswig, M.: Polymake: a framework for analyzing convex polytopes. In: Polytopes—combinatorics and computation. pp. 43–73. Springer (2000), https: //doi.org/10.1007/978-3-0348-8438-9_2

    2000 doi

  10. [16]

    Computer PhysicsCommunications 234,202–215(2019),https://doi.org/10.1016/j.cpc.2018

    Gerdt, V.P., Lange-Hegermann, M., Robertz, D.: The MAPLE package TDDS for computing Thomas decompositions of systems of nonlinear PDEs. Computer PhysicsCommunications 234,202–215(2019),https://doi.org/10.1016/j.cpc.2018. 07.025

    2019 doi

  11. [17]

    In: Proceedings of the 2019 International Symposium on Symbolic and Algebraic Computation

    Gerdt, V.P., Robertz, D.: Algorithmic approach to strong consistency analysis of finite difference approximations to PDE systems. In: Proceedings of the 2019 International Symposium on Symbolic and Algebraic Computation. ISSAC ’19 (Jul 2019), http://dx.doi.org/10.1145/3326229.3326255

    2019

  12. [18]

    Grigor’ev, D.: Complexity of quantifier elimination in the theory of ordinary differ- ential equations, p. 11–25. Springer Berlin Heidelberg (1989), http://dx.doi.org/ 10.1007/3-540-51517-8_81

    1989 doi

  13. [19]

    Journal of Symbolic Computation85, 128–147 (Mar 2018), http://dx.doi.org/10.1016/j.jsc.2017.07.006

    Gustavson, R., Ovchinnikov, A., Pogudin, G.: New order bounds in differential elimination algorithms. Journal of Symbolic Computation85, 128–147 (Mar 2018), http://dx.doi.org/10.1016/j.jsc.2017.07.006

    2018 doi

  14. [21]

    Journal of Symbolic Computation 29(4–5), 641–662 (May 2000), http://dx.doi

    Hubert, E.: Factorization-free decomposition algorithms in differential algebra. Journal of Symbolic Computation 29(4–5), 641–662 (May 2000), http://dx.doi. org/10.1006/jsco.1999.0344

  15. [22]

    In: Lecture Notes in Computer Science, pp

    Hubert, E.: Notes on triangular sets and triangulation-decomposition algorithms II: Differential systems. In: Lecture Notes in Computer Science, pp. 40–87. Springer Berlin Heidelberg (2003), https://doi.org/10.1007/3-540-45084-x_2

    2003 doi

  16. [23]

    Mukhina and G

    Kauers, M., Pages, R.: Bounds for D-algebraic closure properties (2025), https: //arxiv.org/abs/2505.07304 20 Y. Mukhina and G. Pogudin

    2025 arXiv

  17. [24]

    Foundations of Computational Mathematics15(2), 451–517 (Feb 2015), http://dx.doi.org/10.1007/s10208-015-9249-9

    Li, W., Yuan, C.M., Gao, X.S.: Sparse differential resultant for Laurent differen- tial polynomials. Foundations of Computational Mathematics15(2), 451–517 (Feb 2015), http://dx.doi.org/10.1007/s10208-015-9249-9

    2015 doi

  18. [25]

    Mathematical Bio- sciences 239(1), 117–123 (Sep 2012), http://dx.doi.org/10.1016/j.mbs.2012.04.008

    Meshkat, N., Anderson, C., DiStefano III, J.J.: Alternative to Ritt’s pseudodivision for finding the input-output equations of multi-output models. Mathematical Bio- sciences 239(1), 117–123 (Sep 2012), http://dx.doi.org/10.1016/j.mbs.2012.04.008

    2012 doi

  19. [26]

    Mohr, R., Mukhina, Y.: On the computation of Newton polytopes of eliminants (2025), https://arxiv.org/abs/2502.05015

    2025 arXiv

  20. [27]

    17–21 (1990), http://dx.doi.org/10.1016/ B978-0-08-037022-4.50009-0

    Moog, C., Perraud, J., Bentz, P., Vo, Q.: Prime differential ideals in non- linear rational controls systems, p. 17–21 (1990), http://dx.doi.org/10.1016/ B978-0-08-037022-4.50009-0

    1990

  21. [28]

    Applied Mathematics and Computation 24(2), 101–113 (1987)

    Morgan, A., Sommese, A.: A homotopy for solving general polynomial systems that respects m-homogeneous structures. Applied Mathematics and Computation 24(2), 101–113 (1987)

    1987

  22. [30]

    Oscar – open source computer algebra research system, version 1.0.0 (2024), https: //www.oscar-system.org

    2024

  23. [31]

    Pascadi, A.: Computer-assisted proofs of congruences for multipartitions and divi- sorfunctionconvolutions,basedonmethodsofdifferentialalgebra.TheRamanujan Journal 57(1), 1–36 (Nov 2021), http://dx.doi.org/10.1007/s11139-021-00506-8

    2021 doi

  24. [32]

    polytechnique.fr/Labo/Gleb.POGUDIN/files/da_notes.pdf

    Pogudin, G.: Lecture notes on differential algebra (2023), http://www.lix. polytechnique.fr/Labo/Gleb.POGUDIN/files/da_notes.pdf

    2023

  25. [33]

    American Mathe- matical Society (1932), https://archive.org/details/differentialequa033050mbp

    Ritt, J.F.: Differential equations from the algebraic standpoint. American Mathe- matical Society (1932), https://archive.org/details/differentialequa033050mbp

    1932

  26. [34]

    Springer International Publishing (2014), http://dx.doi.org/10.1007/978-3-319-11445-3

    Robertz, D.: Formal Algorithmic Elimination for PDEs. Springer International Publishing (2014), http://dx.doi.org/10.1007/978-3-319-11445-3

    2014 doi

  27. [35]

    Advances in Applied Mathematics72, 4–37 (Jan 2016), http://dx.doi

    Rueda, S.L.: Differential elimination by differential specialization of Sylvester style matrices. Advances in Applied Mathematics72, 4–37 (Jan 2016), http://dx.doi. org/10.1016/j.aam.2015.07.002

    2016 doi

  28. [36]

    Shafarevich, I.R., Reid, M.: Basic algebraic geometry, vol. 2. Springer (1994)

    1994

  29. [37]

    Simmons, W., Platzer, A.: Differential elimination and algebraic invariants of poly- nomial dynamical systems (2023), https://arxiv.org/abs/2301.10935

    2023

  30. [38]

    Journal of Symbolic Computation 126, 102348 (Jan 2025), http://dx.doi.org/10.1016/j.jsc.2024.102348

    Teguia Tabuguia, B.: Arithmetic of D-algebraic functions. Journal of Symbolic Computation 126, 102348 (Jan 2025), http://dx.doi.org/10.1016/j.jsc.2024.102348

    2025

  31. [39]

    In: Mathematical Software (2002), https://doi.org/10.1142/9789812777171_0040

    Wang, D.: EPSILON: A library of software tools for polynomial elimination. In: Mathematical Software (2002), https://doi.org/10.1142/9789812777171_0040

    2002 doi

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.