https://github.com/keisukenagakawa/cisim

Hypegeometric confidence interval calculator.
https://github.com/keisukenagakawa/cisim

scipy simulation statistics

Last synced: 27 days ago
JSON representation

Hypegeometric confidence interval calculator.

• Host: GitHub
• URL: https://github.com/keisukenagakawa/cisim
• Owner: KeisukeNagakawa
• License: mit
• Created: 2020-07-15T12:29:34.000Z (over 4 years ago)
• Default Branch: master
• Last Pushed: 2020-10-19T12:04:35.000Z (about 4 years ago)
• Last Synced: 2024-10-11T12:21:48.828Z (27 days ago)
• Topics: scipy, simulation, statistics
• Language: Python
• Homepage:
• Size: 58.6 KB
• Stars: 2
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.rst
  • License: LICENSE

Awesome Lists containing this project

README

        
Confidence Interval of Hypergeometric Distribution Simulation Module

=====================================================================

For a hypergeometric distribution, denoted by :math:`Hyper(n_pop, k_s, n_draw)`, where :math:`n_pop` is the population size, :math:`k_s` is the number of population with 'success', and :math:`n_draw` is the given sample size, this program calculates the confidence interval for :math:`k_s` when :math:`n_pop` is known and number of observed success is given as :math:`k_s_obs`, using simulation scipy optimization tools.

Note that this is done by a simulation and can have unaccurate values as a solution. Please cross-check with other site calculating confidence interval of hypergeometric function like here:  `cluster-text.com `_.

PyPI package link is `here `_.

===============================================================

First, pip install this package.

shell

.. code-block:: shell

    pip install cisim

Note: if this installing doesn't go well, please try pip install these packages: cerberus, numpy, and scipy.

After the pip-install, you can calculate the interval of a Hypergeometric distribution as follows.

python

.. code-block:: python3

    import cisim

    from cisim.stats import HyperCI

    h = HyperCI(n_pop=10**4, n_draw=10**3, k_s_obs=100)

    res = h.ci_sim()

    print(res[0])

    >>> [830, 1193]

And this program can also calculate confidence interaval for binomial distribution, too.

.. code-block:: python3

    import cisim

    from cisim.stats import BinomCI

    b = BinomCI(n_pop=100, n_obs=10, cl=0.05)

    res = b.ci_sim()

    print(res[0])

    >>> [0.049005430267763495, 0.17622473596973592]

You can check and compare the result with Clopper-Pearson approach on `cluster-text.com `_

Learn more at `HyperGeometric distribution `_, `Binomial Distribution `_

---------------

If you want to learn more about ``setup.py`` files, check out `this repository `_.