estimation.feglm_.Feglm

estimation.feglm_.Feglm(
    self,
    FixestFormula,
    data,
    ssc_dict,
    drop_singletons,
    drop_intercept,
    weights,
    weights_type,
    collin_tol,
    fixef_tol,
    lookup_demeaned_data,
    tol,
    maxiter,
    solver,
    store_data=True,
    copy_data=True,
    lean=False,
    sample_split_var=None,
    sample_split_value=None,
    separation_check=None,
    context=0,
)

Abstract base class for the estimation of a fixed-effects GLM model.

Methods

Name	Description
get_fit	Fit the GLM model via iterated weighted least squares.
predict	Return predicted values from regression model.
prepare_model_matrix	Prepare model inputs for estimation.
residualize	Residualize v and X by flist using weights.
to_array	Turn estimation DataFrames to np arrays.

get_fit

estimation.feglm_.Feglm.get_fit()

Fit the GLM model via iterated weighted least squares.

All equations in the code refer to Stammann ‘Fast and Feasible Estimation of Generalized Linear Models with High-Dimensional k-way fixed effects’: https://arxiv.org/pdf/1707.01815

predict

estimation.feglm_.Feglm.predict(
    newdata=None,
    atol=1e-06,
    btol=1e-06,
    type='link',
    se_fit=False,
    interval=None,
    alpha=0.05,
)

Return predicted values from regression model.

Return a flat np.array with predicted values of the regression model. If new fixed effect levels are introduced in newdata, predicted values for such observations will be set to NaN.

Parameters

Name	Type	Description	Default
newdata	Union[None, pd.DataFrame]	A pd.DataFrame with the new data, to be used for prediction. If None (default), uses the data used for fitting the model.	None
atol	Float	Stopping tolerance for scipy.sparse.linalg.lsqr(). See https://docs.scipy.org/doc/ scipy/reference/generated/scipy.sparse.linalg.lsqr.html	1e-6
btol	Float	Another stopping tolerance for scipy.sparse.linalg.lsqr(). See https://docs.scipy.org/doc/ scipy/reference/generated/scipy.sparse.linalg.lsqr.html	1e-6
type	str	The type of prediction to be computed. Can be either “response” (default) or “link”. If type=“response”, the output is at the level of the response variable, i.e., it is the expected predictor E(Y|X). If “link”, the output is at the level of the explanatory variables, i.e., the linear predictor X @ beta.	'link'
atol	Float	Stopping tolerance for scipy.sparse.linalg.lsqr(). See https://docs.scipy.org/doc/scipy/reference/generated/scipy.sparse.linalg.lsqr.html	1e-6
btol	Float	Another stopping tolerance for scipy.sparse.linalg.lsqr(). See https://docs.scipy.org/doc/scipy/reference/generated/scipy.sparse.linalg.lsqr.html	1e-6
se_fit	Optional[bool]	If True, the standard error of the prediction is computed. Only feasible for models without fixed effects. GLMs are not supported. Defaults to False.	False
interval	Optional[PredictionErrorOptions]	The type of interval to compute. Can be either ‘prediction’ or None.	None
alpha	float	The alpha level for the confidence interval. Defaults to 0.05. Only used if interval = “prediction” is not None.	0.05

Returns

Name	Type	Description
	Union[np.ndarray, pd.DataFrame]	Returns a pd.Dataframe with columns “fit”, “se_fit” and CIs if argument “interval=prediction”. Otherwise, returns a np.ndarray with the predicted values of the model or the prediction standard errors if argument “se_fit=True”.

prepare_model_matrix

estimation.feglm_.Feglm.prepare_model_matrix()

Prepare model inputs for estimation.

residualize

estimation.feglm_.Feglm.residualize(v, X, flist, weights, tol)

Residualize v and X by flist using weights.

to_array

estimation.feglm_.Feglm.to_array()

Turn estimation DataFrames to np arrays.