HN Showcase
Back to browse
GitHub Repository

Evaluate Python functions at points where they're undefined.

6 starsPython

Evaluate Python functions at their singularities

by calculusmachine·Apr 15, 2026·1 point·0 comments
Visit ProjectView on HN

AI Analysis

●●●BangerBig BrainNiche GemWizardry

First library to resolve Python function singularities algebraically without symbolic expressions.

Strengths
  • •@safe decorator adds zero overhead to normal execution, only resolves on failure
  • •Classifies singularities as Removable, Pole, or Regular with provenance tracking
  • •Pure Python with zero dependencies, works with standard math module functions
Weaknesses
  • •Niche audience limits adoption — most Python devs never hit singularities
  • •No documentation on performance characteristics for large-scale numerical workloads
Target Audience

Scientific computing developers, numerical analysts, math-heavy Python users

Similar To

SymPy · mpmath

Post Description

The library where you pass a plain numeric Python function and get exact limits via algebraic infinitesimal arithmetic.

To my knowledge, this is the first library that does this directly on plain Python functions.

It eliminates edge-case handling and numerical instability at singularities, you just write the function and evaluate it everywhere.

Similar Projects