segreg.bootstrap.model_bca

model_bca(boot_sims, orig_sample_estimate, estimator, indep, dep, significance=0.05, no_acceleration=False, verbose=False)[source]

Computes BCa (bias-corrected and accelerated) confidence intervals for regression models.

Note

This computes BCa confidence intervals for each parameter independently. In particular, it does not compute any joint parameter region.

Warning

If there are any fixed parameters in the model, they are assumed to be excluded from the arguments passed to this method.

Parameters

  • boot_sims (numpy ndarray of shape: (num_simulations, num_params)) – Each column represents bootstrap simulations for a single statistic. Should not include any columns corresponding to fixed parameters in the model.

  • orig_sample_estimate (numpy ndarray shape: (num_params,)) – The estimate for the parameters from the original data sample. It should not include fixed parameters.

  • estimator (subclass of segreg.model.estimator.Estimator) –

  • indep (array-like) – The independent data. Also called predictor, explanatory variable, regressor, or exogenous variable.

  • dep (array-like) – The dependent data. Also called response, regressand, or endogenous variable.

Returns

bca confidence intervals – Columns are endpoints of confidence intervals, left, right, respectively.

Return type

numpy array of shape (num_boot_params, 2)