estimation.model_matrix_fixest

estimation.model_matrix_fixest(FixestFormula, data, drop_singletons=False, weights=None, drop_intercept=False)

Create model matrices for fixed effects estimation.

This function processes the data and then calls formulaic.Formula.get_model_matrix() to create the model matrices.

Parameters

Name	Type	Description	Default
`FixestFormula`	A pyfixest.estimation.FormulaParser.FixestFormula object	that contains information on the model formula, the formula of the first and second stage, dependent variable, covariates, fixed effects, endogenous variables (if any), and instruments (if any).	required
`data`	pd.DataFrame	The input DataFrame containing the data.	required
`drop_singletons`	bool	Whether to drop singleton fixed effects. Default is False.	`False`
`weights`	str or None	A string specifying the name of the weights column in `data`. Default is None.	`None`
`data`	pd.DataFrame	The input DataFrame containing the data.	required
`drop_intercept`	bool	Whether to drop the intercept from the model matrix. Default is False. If True, the intercept is dropped ex post from the model matrix created by formulaic.	`False`

Returns

Type	Description
dict	A dictionary with the following keys and value types: - ‘Y’ : pd.DataFrame The dependent variable. - ‘X’ : pd.DataFrame The Design Matrix. - ‘fe’ : Optional[pd.DataFrame] The model’s fixed effects. None if not applicable. - ‘endogvar’ : Optional[pd.DataFrame] The model’s endogenous variable(s), None if not applicable. - ‘Z’ : np.ndarray The model’s set of instruments (exogenous covariates plus instruments). None if not applicable. - ‘weights_df’ : Optional[pd.DataFrame] DataFrame containing weights, None if weights are not used. - ‘na_index’ : np.ndarray Array indicating rows droppled beause of NA values or singleton fixed effects. - ‘na_index_str’ : str String representation of ‘na_index’. - ’_icovars’ : Optional[list[str]] List of variables interacted with i() syntax, None if not applicable. - ‘X_is_empty’ : bool Flag indicating whether X is empty.

dict

A dictionary with the following keys and value types: - ‘Y’ : pd.DataFrame The dependent variable. - ‘X’ : pd.DataFrame The Design Matrix. - ‘fe’ : Optional[pd.DataFrame] The model’s fixed effects. None if not applicable. - ‘endogvar’ : Optional[pd.DataFrame] The model’s endogenous variable(s), None if not applicable. - ‘Z’ : np.ndarray The model’s set of instruments (exogenous covariates plus instruments). None if not applicable. - ‘weights_df’ : Optional[pd.DataFrame] DataFrame containing weights, None if weights are not used. - ‘na_index’ : np.ndarray Array indicating rows droppled beause of NA values or singleton fixed effects. - ‘na_index_str’ : str String representation of ‘na_index’. - ’_icovars’ : Optional[list[str]] List of variables interacted with i() syntax, None if not applicable. - ‘X_is_empty’ : bool Flag indicating whether X is empty.