QuantregSolution

QuantregSolution(beta, has_converged, iterations, x, s, z, w, y)

Solution of the Frisch-Newton interior point solver.

Fields follow the notation of Portnoy and Koenker (1997), Statistical Science.

Parameters

Name Type Description Default
beta NDArray[np.float64] Coefficient estimates, shape (n_coefficients,); the negated y. required
has_converged bool Whether the duality gap fell below the tolerance within the iteration budget. required
iterations int Index of the final interior point iteration, counted from zero. required
x NDArray[np.float64] Primal variable, shape (n_solver_rows,). required
s NDArray[np.float64] Primal slack of the upper bound, u - x, shape (n_solver_rows,). required
z NDArray[np.float64] Dual variable of x >= 0, shape (n_solver_rows,). required
w NDArray[np.float64] Dual variable of x <= u, shape (n_solver_rows,). required
y NDArray[np.float64] Dual variable of A x = b, shape (n_coefficients,). required

Examples

import pyfixest as pf

fit = pf.quantreg("Y ~ X1", pf.get_data(), quantile=0.5)
fit.solution.has_converged, fit.solution.iterations
(True, 10)