∇ py-nabla: Mathematics in the language of thought.

py-nabla is a production-grade mathematical computing engine that bridges the gap between LaTeX notation and Python execution. Write math exactly as you think it; execute it with the power of SymPy and NumPy.

🔥 Why py-nabla?

Translating complex LaTeX equations into nested Python code is brittle, time-consuming, and error-prone. One misplaced parenthesis in a symbolic expression can lead to hours of debugging.

py-nabla eliminates this friction.

• Parse Naturally: High-fidelity Earley parser for complex LaTeX (vmatrix, cases, summations, mixed partials).
• Compute Symbolically: Full integration with SymPy for calculus, algebra, and simplification.
• Execute Numerically: Instant vectorization via NumPy for high-performance evaluation.
• Visualize Beautifully: Built-in 2D/3D plotting engine with publication-ready aesthetics.
• Render Professionally: Bidirectional LaTeX conversion with support for Leibniz, Prime, and Newton (dot) notations.

🛠️ Installation

# Core engine
pip install py-nabla

# With plotting support (recommended)
pip install py-nabla[plotting]

# Install all development dependencies
pip install py-nabla[all]

✨ Quickstart: The "Wow" Moment

Differentiate a complex expression, simplify it, and plot it in seconds:

import py_nabla as nb
import numpy as np

# 1. Parse your LaTeX
f = nb.parse(r"\frac{d}{dx} \left( x^2 \sin(x) \right)")

# 2. Compute analytically
print(f"Derivative: {f.simplify()}") 
# Result: 2*x*sin(x) + x**2*cos(x)

# 3. Render back to LaTeX (Leibniz style)
print(f.latex(mode='display'))

# 4. Solve Differential Equations (v1.1.0+)
# Solve y'' + y = 0 natively
eq = nb.parse(r"y'' + y = 0")
print(f"Solution: {eq.dsolve()}") # y(t) = C1*sin(t) + C2*cos(t)

# 5. Plot instantly
nb.plot(f, domain=(-2, 2), title="Differentiated Waveform")

💎 Production Features (v1.1.0)

🧠 Advanced Parser

• Calculus: Support for $\int, \iint, \oint$, $\lim_{x \to a}$, $\sum, \prod$, and partial derivatives.
• Differential/Integral Equations: Solve ODEs, IDEs, and Fractional Differential Equations (FDEs) using dsolve() with a built-in Laplace Transform engine. Supports Newton's prime notation ($y', y''$) and fractional orders ($y^{(0.5)}$) natively.
• Linear Algebra: Matrix environments (bmatrix, pmatrix, vmatrix) and vector notations.
• Piecewise: Support for \begin{cases} environments.
• Ambiguity Resolution: Intelligent splitting of multi-char symbols (e.g., 2weight $\to 2 \cdot w \cdot e \cdot i \cdot g \cdot h \cdot t$).

🪄 Publication Rendering

Beautiful bidirectional conversion with custom styles:

• Derivative Styles: Choose between leibniz ($\frac{dy}{dx}$), prime ($y'$), or dot ($\dot{y}$).
• Multiple Modes: inline, display, equation, or align.

📊 Plotting Engine

• 2D Plots: Multi-function plotting with automatic labeling.
• 3D Plots: Surface, wireframe, and contour plots for multivariable functions.
• Parametric: Plot complex curves like circles, spirals, or Lissajous figures.

📚 Documentation & Tutorials

• Getting Started Guide
• Calculus Cookbook
• Linear Algebra Mastery
• Advanced Plotting
• Differential & Integral Equations
• Fractional Calculus Guide

🤝 Contributing

We welcome contributions from the mathematical and open-source communities! Please see CONTRIBUTING.md for guidelines, and our Testing Guide before submitting code.

📄 License

py-nabla is released under the MIT License.