FixestMulti

FixestMulti(config, parsed, data, context)

Results container holding every model fitted by one public-API call.

Returned instead of a single model when the formula uses multiple estimation syntax: several dependent variables, the stepwise operators (sw, sw0, csw, csw0, mvsw), stepwise fixed effects, or a split / fsplit sample split. Reporting methods such as etable(), summary() and coefplot() apply to all fitted models at once. fetch_model() returns an individual model.

Examples

import pyfixest as pf

data = pf.get_data()
fits = pf.feols("Y ~ X1 | csw0(f1, f2)", data)

pf.etable(fits)

	Y
	(1)	(2)	(3)
coef
X1	-1.000*** (0.085)	-0.949*** (0.070)	-0.919*** (0.060)
Intercept	0.919*** (0.112)
fe
f2	-	-	x
f1	-	x	-
f1	-	-	x
stats
Observations	998	997	997
R2	0.123	0.437	0.609
Significance levels: * p < 0.05, ** p < 0.01, *** p < 0.001. Format of coefficient cell: Coefficient (Std. Error)

Retrieve one of the fitted models to work with it on its own:

fit = fits.fetch_model(0)
fit.tidy()

Model:  Y ~ X1

	Estimate	Std. Error	t value	Pr(>|t|)	2.5%	97.5%
Coefficient
Intercept	0.918518	0.111821	8.214166	6.661338e-16	0.699086	1.137950
X1	-1.000086	0.084735	-11.802468	0.000000e+00	-1.166366	-0.833806

Attributes

Name	Description
FixestMulti.FixestFormulaDict	Parsed formula dict keyed by fixed-effects spec.

Methods

Name	Description
FixestMulti.confint	Obtain confidence intervals for the fitted models.
FixestMulti.fetch_model	Fetch a model of class Feols from the Fixest class.
FixestMulti.tidy	Return the results of an estimation using feols() as a tidy Pandas DataFrame.
FixestMulti.to_list	Return a list of all fitted models.
FixestMulti.vcov	Update regression inference “on the fly”.
FixestMulti.wildboottest	Run a wild cluster bootstrap for all regressions in the Fixest object.

FixestMulti.confint

confint()

Obtain confidence intervals for the fitted models.

Returns

Name	Type	Description
	pandas.Series	A pd.Series with coefficient names and confidence intervals. The key indicates which models the estimated statistic derives from.

FixestMulti.fetch_model

fetch_model(i, print_fml=True)

Fetch a model of class Feols from the Fixest class.

Parameters

Name	Type	Description	Default
i	int or str	The index of the model to fetch.	required
print_fml	bool	Whether to print the formula of the model. Default is True.	True

Returns

Name	Type	Description
	A Feols object.

FixestMulti.tidy

tidy()

Return the results of an estimation using feols() as a tidy Pandas DataFrame.

Returns

Name	Type	Description
	pandas.DataFrame or str	A tidy DataFrame with the following columns: - fml: the formula used to generate the results - Coefficient: the names of the coefficients - Estimate: the estimated coefficients - Std. Error: the standard errors of the estimated coefficients - t value: the t-values of the estimated coefficients - Pr(>|t|): the p-values of the estimated coefficients - 2.5%: the lower bound of the 95% confidence interval - 97.5%: the upper bound of the 95% confidence interval If type is set to “markdown”, the resulting DataFrame will be returned as a markdown-formatted string with three decimal places.

FixestMulti.to_list

to_list()

Return a list of all fitted models.

Returns

Name	Type	Description
	list	A list of all fitted models of types Feols, Fepois, or Feiv.

FixestMulti.vcov

vcov(vcov, vcov_kwargs=None)

Update regression inference “on the fly”.

By calling vcov() on a “Fixest” object, all inference procedures applied to the “Fixest” object are replaced with the variance-covariance matrix specified via the method.

Parameters

Name	Type	Description	Default
vcov	Union[str, dict[str, str]])	A string or dictionary specifying the type of variance-covariance matrix to use for inference. - If a string, can be one of “iid”, “hetero”, “HC1”, “HC2”, “HC3”. - If a dictionary, it should have the format {“CRV1”: “clustervar”} for CRV1 inference or {“CRV3”: “clustervar”} for CRV3 inference.	required
vcov_kwargs	Optional[dict[str, any]]	Additional keyword arguments for the variance-covariance matrix.	None

Returns

Name	Type	Description
	An instance of the "Fixest" class with updated inference.

FixestMulti.wildboottest

wildboottest(
    reps,
    cluster=None,
    param=None,
    weights_type='rademacher',
    impose_null=True,
    bootstrap_type='11',
    seed=None,
    k_adj=True,
    G_adj=True,
)

Run a wild cluster bootstrap for all regressions in the Fixest object.

Parameters

Name	Type	Description	Default
B	int	The number of bootstrap iterations to run.	required
param	Union[str, None]	A string of length one, containing the test parameter of interest. Default is None.	None
cluster	Union[str, None]	The name of the cluster variable. Default is None. If None, uses the self._clustervar attribute as the cluster variable. If the self._clustervar attribute is None, a heteroskedasticity-robust wild bootstrap is run.	None
weights_type	str	The type of bootstrap weights. Either ‘rademacher’, ‘mammen’, ‘webb’, or ‘normal’. Default is ‘rademacher’.	'rademacher'
impose_null	bool	Should the null hypothesis be imposed on the bootstrap dgp, or not? Default is True.	True
bootstrap_type	str	A string of length one. Allows choosing the bootstrap type to be run. Either ‘11’, ‘31’, ‘13’, or ‘33’. Default is ‘11’.	'11'
seed	Union[str, None]	Option to provide a random seed. Default is None.	None
k_adj	bool	Whether to adjust the original coefficients with the bootstrap distribution. Default is True.	True
G_adj	bool	Whether to adjust standard errors for clustering in the bootstrap. Default is True.	True

Returns

Name	Type	Description
	pandas.DataFrame	A pd.DataFrame with bootstrapped t-statistic and p-value. The index indicates which model the estimated statistic derives from.