API

segreg.model: Models

OLSRegressionEstimator()	Estimator for ordinary least-squares regression.
OneBkptSegRegEstimator([num_end_to_skip, ...])	Estimator for one-bkpt segmented regression.
TwoBkptSegRegEstimator([num_end_to_skip, ...])	Estimator for two-bkpt segmented regression.
one_bkpt_rss_func(indep, dep)	Returns a function which takes a bkpt and returns the RSS of a one-bkpt segmented fit.
two_bkpt_rss_func(indep, dep)	Returns a function which takes a pair of bkpts and returns the RSS of a two-bkpt segmented fit.

segreg.analysis: Tools for Analyzing Results

plot_models(func_arr, indep, dep[, ...])	Plots univariate functions together with scatter of the data.
plot_one_bkpt_segreg_rss(indep, dep[, lines])
plot_two_bkpt_segreg_rss(indep, dep)

segreg.bootstrap: Bootstrap

boot_conf_intervals(indep, dep, estimator[, ...])	Bootstrap confidence intervals for regression models.
bca_acceleration(estimator, indep, dep)	see: stata manual on the bootstrap
model_bca(boot_sims, orig_sample_estimate, ...)	Computes BCa (bias-corrected and accelerated) confidence intervals for regression models.
bca(boot_sims, orig_sample_estimate, ...[, ...])	Computes BCa (bias-corrected and accelerated) confidence intervals for a single parameter.
boot_basic_conf_interval(boot_sims, ...[, ...])	Name follows Davison and Hinkley.
boot_percentile_conf_interval(boot_sims, ...)
boot_param_dist(indep, dep, estimator, num_sims)	Computes bootstrap sampling distribution of model parameters.
boot_resample(indep, dep[, fitted_values, ...])	A single boot resampling for a regression model.
random_selection_with_replacement(series)	Random draw with replacement from series.
random_selection_with_replacement_two_series(...)	Random draw with replacement.

segreg.data: Datasets

hinkley()	A dataset from an academic paper by David Hinkley.
test1()	Random dataset for testing.
test2()	Random dataset for testing.

segreg.model.alt: Alternative Implementations

Pure Python Segmented Regression

fit_one_bkpt(indep, dep[, num_end_to_skip, ...])	Estimate one-bkpt segmented regression model.
fit_two_bkpt(indep, dep[, num_end_to_skip, ...])	Estimate two-bkpt segmented regression model.

Brute Force Grid Search

brute_fit_one_bkpt(indep, dep[, ...])	Estimate one-bkpt segmented regression model using a brute-force method.
brute_fit_two_bkpt(indep, dep[, ...])	Estimate two-bkpt segmented regression model using a brute-force method.