ebcic

EBCIC: Exact Binomial Confidence Interval Calculator

---

Project maintained by KazKobara Hosted on GitHub Pages — Theme by mattgraham

Change Log

All notable changes, such as backward incompatibilities, will be documented in this file.

[0.0.4]

Added

• Rejection area parameters, rej_perc_upper, rej_perc_lower, rej_upper and rej_lower to set and manage confidence interval.
• Doctest and unittest in test.py.

Changed

• Moved main() to cli() to enable tests.
• Moved alpha_to_za() and alpha_to_zah() from Params into normal_approx(), wilson_score() and wilson_score_cc().
• Enabled Params.rule_of_ln_alpha(), Params.beta_approx(), Params.normal_approx(), Params.wilson_score(), Params.wilson_score_cc(), Params.verify_interval_of_p() and Params.print_interval().
• Sphinx 5.1.1.

[0.0.3]

Added

• Approximation using the beta distribution.
• Params.exact_border() and Params.exact().

Changed

• Max of x-axis in a graph to (k_end * 10**log_n_end) from (3 * k_end * 10**log_n_end).
• Made plot settings in interval_graph() in common as set_graph().
• For k=0, interval_graph() shows only upper intervals due to set_graph().
• Made k=0 and k=n be one-sided for ‘wilson’, ‘wilson_cc’ and ‘beta_approx’ in accordance with ‘exact’.

[0.0.2]

Added

• Usage in MATLAB.
• Notes that ‘confidence percentage’ of k=0 or k=n is for one-sided.

[0.0.1]

Added

• Command line program using ‘ebcic’ package.
• ‘ebcic’ package is available with pip install ebcic