How pyportfolio/pyportfolioopt works
PyPortfolioOpt is a Python library for estimating portfolio inputs, optimizing continuous asset weights, and converting weights into purchasable share allocations. Its principal workflow, documented in the README, starts with price data, estimates expected returns and risk, optimizes weights, then optionally discretizes them using latest prices and a portfolio budget. The library also offers Black–Litterman, Critical Line Algorithm, hierarchical risk parity, and alternative risk objectives. Internal optimizer and objective-function modules support several optimization paths. Some subsystems are represented from the README and source index; unsampled implementation details and links are omitted where their runtime relationships are not shown.
Main entry points: efficient_frontier.py Mean-variance frontier · _base_optimizer.py Optimizer base · risk_models.py Risk models · discrete_allocation.py Share allocation
Components 19
Return And Risk Inputs
- Expected returns
pypfopt/expected_returns.py - Risk models
pypfopt/risk_models.py
Portfolio Optimization
- Mean-variance frontier
pypfopt/efficient_frontier/efficient_frontier.py - Critical Line Algorithm
pypfopt/cla.py - Black-Litterman
pypfopt/black_litterman.py - Hierarchical Risk Parity
pypfopt/hierarchical_portfolio.py - Semivariance frontier
pypfopt/efficient_frontier/efficient_semivariance.py - CVaR frontier
pypfopt/efficient_frontier/efficient_cvar.py - CDaR frontier
pypfopt/efficient_frontier/efficient_cdar.py
Portfolio Construction
- Share allocation
pypfopt/discrete_allocation.py - Portfolio plotting
pypfopt/plotting.py
Optimizer Foundation
- Objective functions
pypfopt/objective_functions.py - Optimizer base
pypfopt/base/_base_optimizer.py - Optimization errors
pypfopt/exceptions.py
Other components
- Portfolio user
- Historical price data
- Continuous portfolio weights
- Latest asset prices
- Discrete holdings
Connections 18
- Portfolio user → Historical price data: supplies
README.md - Historical price data → Expected returns: estimates returns
README.md - Historical price data → Risk models: estimates covariance
README.md - Expected returns → Mean-variance frontier: provides estimates
README.md - Risk models → Mean-variance frontier: provides covariance
README.md - Mean-variance frontier → Continuous portfolio weights: optimizes weights
README.md - Continuous portfolio weights → Share allocation: converts weights
README.md - Latest asset prices → Share allocation: prices assets
README.md - Share allocation → Discrete holdings: returns holdings
README.md - Mean-variance frontier → Optimizer base: inherits optimizer
pypfopt/efficient_frontier/efficient_frontier.py - Critical Line Algorithm → Optimizer base: inherits optimizer
pypfopt/cla.py - Black-Litterman → Optimizer base: inherits optimizer
pypfopt/black_litterman.py - Semivariance frontier → Mean-variance frontier: extends frontier
pypfopt/efficient_frontier/efficient_semivariance.py - CVaR frontier → Mean-variance frontier: extends frontier
pypfopt/efficient_frontier/efficient_cvar.py - CDaR frontier → Mean-variance frontier: extends frontier
pypfopt/efficient_frontier/efficient_cdar.py - Hierarchical Risk Parity → Risk models: uses risk models
pypfopt/hierarchical_portfolio.py - Mean-variance frontier → Objective functions: uses objectives
pypfopt/efficient_frontier/efficient_frontier.py - Optimizer base → Optimization errors: raises errors
pypfopt/base/_base_optimizer.py