models

A namespace of namespaces over portfolio construction, optimization, and management. See Construction, optimization & management for a worked walkthrough, or examples/models_debug.ipynb/examples/models_tutorial.ipynb for runnable, end-to-end tours.

Submodule	Covers
base	ModelResult, constraint objects (WeightBounds, GroupCap, TurnoverCap, NetExposure, GrossExposure), and the shared multi-start SLSQP solve routine.
estimators	Expected returns (mean-historical/EWMA/CAPM-implied/James-Stein), covariance (sample/EWMA/shrinkage/Ledoit-Wolf/robust), and factor models (CAPM, Fama-French, rolling/linear regression, factor attribution).
optimization	Pure solver functions: mean-variance, max Sharpe, min variance, target return/volatility, efficient frontier, Black-Litterman, hierarchical risk parity, risk parity/budgeting, maximum diversification.
construction	Constraint objects (re-exported from base), plus optimize/build/efficient_frontier - the recipe layer over optimization.
management	rebalance (trade lists), monitor (drift/limit checks), compare (before/after or portfolio-vs-portfolio).

Portfolio.models exposes all of the above bound to a portfolio's own data, with expected_returns/cov_matrix/current_weights/etc. auto-filled - the same _AutoFillNamespace pattern Portfolio.metrics uses, see The Portfolio object.

Not yet implemented: portpy.models.simulation (stochastic processes, Monte Carlo, time series, ML - see the Roadmap).