Feols.update

update(X_new, y_new, inplace=False)

Update coefficients for new observations using Sherman-Morrison formula.

Parameters

Name	Type	Description	Default
X_new	np.ndarray	Covariates for new data points. Users expected to ensure conformability with existing data.	required
y_new	np.ndarray	Outcome values for new data points.	required
inplace	bool	Whether to update the model object in place. Defaults to False.	`False`

Returns

Name	Type	Description
	np.ndarray	Updated coefficients.

Notes

Updates the coefficients in closed form via the Sherman-Morrison identity instead of refitting on the full sample. X_new has to include the intercept column. Models with fixed effects are not supported.

Examples

Fit on all but the last observation, then add it:

import numpy as np
import pyfixest as pf

data = pf.get_data().dropna()
fit = pf.feols("Y ~ X1 + X2", data.iloc[:-1])

last = data.iloc[[-1]]
X_new = np.column_stack(
    [np.ones(1), last["X1"].to_numpy(), last["X2"].to_numpy()]
)
y_new = last["Y"].to_numpy()

fit.update(X_new, y_new)

array([ 0.88955689, -0.99519687, -0.17661729])