Core API

The core module contains the main classes for dynamic beta estimation.

DynamicBeta

DynamicBetaConfig

Example Usage

Basic Usage

from grubeta import DynamicBeta

# Create model with default settings
model = DynamicBeta()

# Or with custom parameters
model = DynamicBeta(
    lookback=60,
    lambda_beta=0.05,
    lambda_alpha=0.5,
    lambda_alpha_smooth=0.1
)

# Fit and predict
results = model.fit_predict(stock_returns, market_returns)

Using Configuration Object

from grubeta import DynamicBeta, DynamicBetaConfig

config = DynamicBetaConfig(
    lookback=90,
    initial_train_size=500,
    wf_step_size=126,
    learning_rate=1e-4,
    lambda_beta=0.05,
    lambda_alpha=0.5,
    lambda_alpha_smooth=0.1
)

model = DynamicBeta(config=config)
results = model.fit_predict(stock_returns, market_returns)

Save and Load

# Save fitted model
model.save('./my_model')

# Load later
model = DynamicBeta.load('./my_model')