Science in Python

NumPy is the array foundation of scientific Python, and nearly everything else builds on its arrays. SciPy adds optimization, integration, interpolation, linear algebra, signal processing, sparse matrices and statistics. When a loop cannot be vectorized, Numba compiles numeric Python functions to machine code through LLVM with a single decorator. SymPy does symbolic mathematics: simplification, calculus and equation solving with exact results. statsmodels covers regression, time series and statistical tests with R-style formulas and detailed summary tables.

Many fields have their own established libraries: Astropy for astronomy, Biopython for biological sequences and structures, RDKit for cheminformatics, ObsPy for seismology, and Pint for physical quantities with units. For modeling, PyMC handles Bayesian probabilistic programming, SimPy discrete-event simulation and Mesa agent-based models. NetworkX analyzes graphs in pure Python, Shapely manipulates planar geometry, and Manim renders mathematical animations.