Proteus Actuarial Library documentation

PAL Open-source actuarial modelling in Python

Build models for the unexpected.

The Proteus Actuarial Library is a fast, lightweight framework for simulation-based actuarial and financial models—with dependency tracking, copulas and optional GPU acceleration built in.

pip install proteusllp-actuarial-library

Start with PAL

Learn the essentials

Create stochastic variables, combine risks and run your first model.

Getting Started with PAL
Model dependence

Apply copulas and understand PAL’s automatic coupling groups.

Coupling Groups, Copulas and Variable Reordering
Price reinsurance

Build XoL layers, towers, reinstatements and aggregate covers.

Pricing an Excess-of-Loss Reinsurance Program
Find a class or method

Browse the complete, searchable reference generated from PAL itself.

API Reference

Why PAL

Fast by default

NumPy and SciPy at the core, with optional CuPy acceleration for large simulation workloads.

Dependence made explicit

Automatic variable coupling and a broad choice of copulas make complex models easier to reason about.

Built for actuarial work

Frequency–severity models, reinsurance contracts, risk measures and capital allocation are first-class concepts.

A small model, end to end

Combine two risks and apply a dependency structure
from pal import copulas, distributions

losses = distributions.Gamma(alpha=2.5, theta=2).generate()
expenses = distributions.LogNormal(mu=1, sigma=0.5).generate()

copulas.GumbelCopula(theta=1.2).apply([losses, expenses])
total = losses + expenses