Skip to content
DraconiticResearch
Search Sources API
197,768 studies31 sourcesUpdated 3 min ago
PreprintarXiv licencearXiv

The local-to-global minimum problem of Calvert and Vamanamurthy: a quartic counterexample in three variables and the cubic case in all dimensions

Abstract

In 1980 Calvert and Vamanamurthy asked whether a polynomial whose only critical point is a local minimum must attain its global minimum there. They gave counterexamples of every degree at least five in two or more variables and settled cubics in two and three variables and quartics in two variables, leaving open quartics in three or more variables and cubics in four or more variables. We resolve both remaining cases. First, we give an explicit quartic in three variables whose unique critical point is a nondegenerate strict local minimum, although the polynomial is unbounded below; the example extends to every dimension at least three, and its critical locus is a single point even over the complex numbers. Second, we prove that in every dimension a polynomial of exact degree three with a local minimum has another critical point, so that the cubic case holds vacuously; the proof is elementary and avoids the topological degree arguments used previously in two and three variables. Together with the results of Calvert and Vamanamurthy, this gives the complete classification of the local-to-global property by degree and dimension.

Related

  1. Optimal central limit theorem for bounded random variables in high dimensions
    P. M. Aronow, Patrick Lopatto
    arXivPreprintarXiv licenceFull textProbabilityStatisticsarXiv
  2. Poisson empirical Bayes estimation of sums of random variables via minimum-distance methods
    Stefano Favaro, Sandra Fortini, Soham Jana
    arXivPreprintCC BY 4.0Full textStatisticsarXiv
  3. A counterexample to Reich's fixed point problem
    Luong Van Nguyen
    arXivPreprintCC BY-NC-ND 4.0Full textGeometry & topologyarXiv
  4. A Counterexample to Talagrand's Operator Cotype Problem
    Xinglong Wu
    arXivPreprintarXiv licenceFull textAnalysis & PDEsProbabilityarXiv
  5. A counterexample to the localization problem for AB rings
    Justin Lyle, Saeed Nasseh
    arXivPreprintarXiv licenceFull textAlgebra & number theoryarXiv
BibTeX
@misc{aduen2026localtoglobal,
  title = {{The local-to-global minimum problem of Calvert and Vamanamurthy: a quartic counterexample in three variables and the cubic case in all dimensions}},
  author = {Hugo Aduén},
  howpublished = {arXiv},
  year = {2026},
  month = nov,
  doi = {10.48550/arxiv.2611.02068},
  eprint = {2611.02068},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2611.02068},
}
RIS
TY  - UNPB
TI  - The local-to-global minimum problem of Calvert and Vamanamurthy: a quartic counterexample in three variables and the cubic case in all dimensions
AU  - Hugo Aduén
PY  - 2026
DA  - 2026/11/01
PB  - arXiv
DO  - 10.48550/arxiv.2611.02068
UR  - https://arxiv.org/abs/2611.02068
AB  - In 1980 Calvert and Vamanamurthy asked whether a polynomial whose only critical point is a local minimum must attain its global minimum there. They gave counterexamples of every degree at least five in two or more variables and settled cubics in two and three variables and quartics in two variables, leaving open quartics in three or more variables and cubics in four or more variables. We resolve both remaining cases. First, we give an explicit quartic in three variables whose unique critical point is a nondegenerate strict local minimum, although the polynomial is unbounded below; the example extends to every dimension at least three, and its critical locus is a single point even over the complex numbers. Second, we prove that in every dimension a polynomial of exact degree three with a local minimum has another critical point, so that the cubic case holds vacuously; the proof is elementary and avoids the topological degree arguments used previously in two and three variables. Together with the results of Calvert and Vamanamurthy, this gives the complete classification of the local-to-global property by degree and dimension.
ER  -