A Python Interface & Extension to Singular

The syngular library is a Python 3 package for algebraic geometry computations. It provides an intuitive and object-oriented interface to Singular. Furthermore, it extends the numerical capabilities of Singular, providing a numerical solver for arbitrary systems of polynomial equations in tandem with pyadic, and its applicaibility to physics computations, where generic algorithms may be insufficient.

Interface

Python classes for 'Ideal', 'Ring' and 'QuotientRing'. Several related functions accessible as attributes or methods. Intuitive operations through magic methods, e.g. Ideal addition '+' and intersection '&'.

Extension

Multivariate solver - i.e. points on varieties over $\mathbb{F}_p$, $\mathbb{Q}_p$ and $\mathbb{C}$.

The function ideal.point_on_variety allows to obtain numerical solutions to arbirary systems of polynomial equations in arbitrary polynomial (quotient) rings, over any of the three above mentioned fields. The system of equations may be underconstrained, i.e. the ideal may have any dimension. The $p$-adic and complex solutions can be requested as not exact, meaning the point may lie close to but not exactly on the associated variety. This is essential for numerical computations where otherwise division-by-zero erros may occur when using exact solutions. The limitation is that Singular must be able to compute an indepednent set for the semi-numerical ideals of low dimension.

Primality test (lighter than a primary decomposition).

The function ideal.test_primality allows to test whether an ideal is prime, primary or neither, without performing a full primary decomposition. The algorithm can run also with successively looser degree bounds. It returns True if the idea is prime, False if it is not, or raises an Inconclusive exception if it cannot decide. If astuple is True, then it will return two booleans: (is_primary, is_prime). Inconclusive cases should now only happen with a Timeout on the computation.

Requirements

numpy, sympy, Singular

Installation

pip install -e path/to/repo

Testing

pytest --cov syngular/ --cov-report html tests/ --verbose

Quick Start

Define an ideal over a ring in two variables

from syngular import Ideal, Ring
I = Ideal(Ring('0', ('x1', 'x2'), 'dp'), ['x1*x2'])

You can now inspect I to see what methods and attributes are available.

Solving arbitrary systems of polynomial equations

Generate a $p$-adic solution to a system of 2 polynomial equations in 3 variables, controlling the precision to which they are solved.

field = Field("padic", 2 ** 31 - 1, 8)
ring = Ring('0', ('x', 'y', 'z', ), 'dp')
I = Ideal(ring, ['x*y^2+y^3-z^2', 'x^3+y^3-z^2', ])

The variety associated to I has 3 branches. In other words, the system of equations has 3 types of solutions.

(Q1, P1), (Q2, P2), (Q3, P3) = I.primary_decomposition

Generate a solution on the first branch

numerical_point = Q1.point_on_variety(field=field, directions=I.generators, valuations=(1, 1, ), ) 

is a dictionary of numerical values for each variable in the ring.

These are small with valuations (1, 1)

list(map(lambda string: Polynomial(string, field).subs(numerical_point), Q1.generators))

while these are O(1) with valuations (0, 0)

list(map(lambda string: Polynomial(string, field).subs(numerical_point), Q2.generators))

See arXiv:2207.10125 Fig. 1 for a graphical depiction.

Citation

If you found this library useful, please consider citing it and Singular

@inproceedings{DeLaurentis:2023qhd,
    author = "De Laurentis, Giuseppe",
    title = "{Lips: $p$-adic and singular phase space}",
    booktitle = "{21th International Workshop on Advanced Computing and Analysis Techniques in Physics Research}: {AI meets Reality}",
    eprint = "2305.14075",
    archivePrefix = "arXiv",
    primaryClass = "hep-th",
    reportNumber = "PSI-PR-23-14",
    month = "5",
    year = "2023"
}

@misc {DGPS,
 title = {{\sc Singular} {4-3-0} --- {A} computer algebra system for polynomial computations},
 author = {Decker, Wolfram and Greuel, Gert-Martin and Pfister, Gerhard and Sch\"onemann, Hans},
 year = {2022},
 howpublished = {\url{http://www.singular.uni-kl.de}},
}