numeraire.baselines.MinVariance#

class numeraire.baselines.MinVariance(*, window: int | None = None, min_obs: int | None = None)[source]#

Bases: object

Global minimum-variance estimator: sample covariance from the (optionally windowed) view.

Parameters:

  • window – Trailing window (in calendar steps) for the sample covariance; None (default) expands from the start. A rolling cap mirrors the skfolio adapter’s estimation window.

  • min_obs – Minimum observations before the first rebalance; None (default) requires strictly more rows than assets, so the sample covariance is non-singular.

__init__(*, window: int | None = None, min_obs: int | None = None) → None[source]#

Methods

__init__(*[, window, min_obs])

fit(view)