Quantreg

Quantreg(
    FixestFormula,
    data,
    ssc_dict,
    drop_singletons,
    drop_intercept,
    weights,
    weights_type,
    collin_tol,
    lookup_demeaned_data,
    solver='np.linalg.solve',
    demeaner=None,
    store_data=True,
    copy_data=True,
    lean=False,
    context=0,
    sample_split_var=None,
    sample_split_value=None,
    quantile=0.5,
    method='fn',
    quantile_tol=1e-06,
    quantile_maxiter=None,
    seed=None,
)

Quantile regression model.

Returned by quantreg(). Fits the conditional quantile of the outcome instead of the conditional mean, which allows the effect of a covariate to differ across the outcome distribution. Estimated via the interior point algorithm of Portnoy and Koenker (1997), Statistical Science.

Examples

import pyfixest as pf

data = pf.get_data()

fit = pf.quantreg("Y ~ X1 + X2", data, quantile=0.5)
fit.tidy()

	Estimate	Std. Error	t value	Pr(>\|t\|)	2.5%	97.5%
Coefficient
Intercept	0.997881	0.172038	5.800355	8.884297e-09	0.660282	1.335480
X1	-1.070846	0.122024	-8.775730	0.000000e+00	-1.310299	-0.831393
X2	-0.182325	0.031911	-5.713491	1.461492e-08	-0.244946	-0.119704

Several quantiles can be estimated in one call. qplot() plots the resulting coefficients.

fits = pf.quantreg("Y ~ X1 + X2", data, quantile=[0.25, 0.5, 0.75])
pf.etable(fits)

	Y
	(1)	(2)	(3)
coef
X1	-0.905*** (0.120)	-1.071*** (0.122)	-1.015*** (0.101)
X2	-0.148*** (0.032)	-0.182*** (0.032)	-0.204*** (0.027)
Intercept	-0.695*** (0.163)	0.998*** (0.172)	2.415*** (0.109)
stats
Observations	998	998	998
R²	-	-	-
Significance levels: * p < 0.05, p < 0.01, * p < 0.001. Format of coefficient cell: Coefficient (Std. Error)

See the quantile regression tutorial for details.

Attributes

Name	Description
Quantreg.objective_value	Compute the total loss of the quantile regression model.

Methods

Name	Description
Quantreg.fit_qreg_fn	Fit a quantile regression model using the Frisch-Newton Interior Point Solver.
Quantreg.fit_qreg_pfn	Fit a quantile regression model using the Frisch-Newton Interior Point Solver with pre-processing.
Quantreg.get_fit	Fit a quantile regression model using the interior point method.
Quantreg.get_performance	Compute performance metrics for the quantile regression model.
Quantreg.prepare_model_matrix	Prepare model inputs for estimation.
Quantreg.to_array	Turn estimation DataFrames to np arrays.

Quantreg.fit_qreg_fn

fit_qreg_fn(X, Y, q, tol=None, maxiter=None, beta_init=None)

Fit a quantile regression model using the Frisch-Newton Interior Point Solver.

Quantreg.fit_qreg_pfn

fit_qreg_pfn(
    X,
    Y,
    q,
    m=None,
    tol=None,
    maxiter=None,
    beta_init=None,
    rng=None,
    eta=None,
)

Fit a quantile regression model using the Frisch-Newton Interior Point Solver with pre-processing.

Quantreg.get_fit

get_fit()

Fit a quantile regression model using the interior point method.

Quantreg.get_performance

get_performance()

Compute performance metrics for the quantile regression model.

Quantreg.prepare_model_matrix

prepare_model_matrix()

Prepare model inputs for estimation.

Quantreg.to_array

to_array()

Turn estimation DataFrames to np arrays.