statsmodels.robust.resistant_linear_model.RLMDetS

class statsmodels.robust.resistant_linear_model.RLMDetS(endog, exog, norm=None, breakdown_point=0.5, col_indices=None, include_endog=False)

S-estimator for linear model with deterministic starts.

Parameters:

endogarray_like, 1-dim

Dependent, endogenous variable.

exogarray_like, 1-dim

Independent, exogenous regressor variables.

normrobust norm

Redescending robust norm used for S-estimation. Default is TukeyBiweight.

breakdown_pointfloat in (0, 0.5)

Breakdown point of the S-estimator.

col_indicesNone or array_like of ints

Index of columns of exog to use in the mahalanobis distance computation for the starting sets of the S-estimator. Default is all exog except first column (constant). Todo: will change when we autodetect the constant column.

include_endogbool

If true, then the endog variable is combined with the exog variables to compute the mahalanobis distances for the starting sets of the S-estimator.

Attributes:

endog_names

Names of endogenous variables

exog_names

Names of exogenous variables

Methods

fit(h[, maxiter, maxiter_step, ...])	Fit a model to data
from_formula(formula, data[, subset, drop_cols])	Create a Model from a formula and dataframe
predict(params[, exog])	After a model has been fit, predict returns the fitted values

Notes

This estimator combines the method of Fast-S regression (Saliban-Barrera et al 2006) using starting sets similar to the deterministic estimation of multivariate location and scatter DetS and DetMM of Hubert et al (2012).

References

[1]

Hubert, Mia, Peter J. Rousseeuw, and Tim Verdonck. 2012. “A Deterministic Algorithm for Robust Location and Scatter.” Journal of Computational and Graphical Statistics 21 (3): 618-37. https://doi.org/10.1080/10618600.2012.672100.

[2]

Hubert, Mia, Peter Rousseeuw, Dina Vanpaemel, and Tim Verdonck. 2015. “The DetS and DetMM Estimators for Multivariate Location and Scatter.” Computational Statistics & Data Analysis 81 (January): 64-75. https://doi.org/10.1016/j.csda.2014.07.013.

[3]

Rousseeuw, Peter J., Stefan Van Aelst, Katrien Van Driessen, and Jose Agulló. 2004. “Robust Multivariate Regression.” Technometrics 46 (3): 293-305.

[4]

Salibian-Barrera, Matías, and Víctor J. Yohai. 2006. “A Fast Algorithm for S-Regression Estimates.” Journal of Computational and Graphical Statistics 15 (2): 414-27.

Methods

fit(h[, maxiter, maxiter_step, ...])	Fit a model to data
from_formula(formula, data[, subset, drop_cols])	Create a Model from a formula and dataframe
predict(params[, exog])	After a model has been fit, predict returns the fitted values

Properties

endog_names	Names of endogenous variables
exog_names	Names of exogenous variables