Covariance Estimator

A covariance estimator estimates the covariance matrix of the assets.

It follows the same API as scikit-learn’s estimator: the fit method takes X as the assets returns and stores the covariance in its covariance_ attribute.

X can be any array-like structure (numpy array, pandas DataFrame, etc.)

Available estimators are:

• EmpiricalCovariance

• EWCovariance

• GerberCovariance

• DenoiseCovariance

• DetoneCovariance

• LedoitWolf

• OAS

• ShrunkCovariance

• GraphicalLassoCV

Example:

from skfolio.datasets import load_sp500_dataset
from skfolio.moments import EmpiricalCovariance
from skfolio.preprocessing import prices_to_returns

prices = load_sp500_dataset()
X = prices_to_returns(prices)

model = EmpiricalCovariance()
model.fit(X)
print(model.covariance_)