{ "info": { "author": "Spencer Beecher", "author_email": "spencebeecher@gmail.com", "bugtrack_url": null, "classifiers": [], "description": "bootstrapped - confidence intervals made easy\n=============================================\n\n**bootstrapped** is a Python library that allows you to build confidence\nintervals from data. This is useful in a variety of contexts - including\nduring ad-hoc a/b test analysis.\n\nMotivating Example - A/B Test\n-----------------------------\n\nImagine we own a website and think changing the color of a 'subscribe'\nbutton will improve signups. One method to measure the improvement is to\nconduct an A/B test where we show 50% of people the old version and 50%\nof the people the new version. We can use the bootstrap to understand\nhow much the button color improves responses and give us the error bars\nassociated with the test - this will give us lower and upper bounds on\nhow good we should expect the change to be!\n\nThe Gist - Mean of a Sample\n---------------------------\n\nGiven a sample of data - we can generate a bunch of new samples by\n're-sampling' from what we have gathered. We calculate the mean for each\ngenerated sample. We can use the means from the generated samples to\nunderstand the variation in the larger population and can construct\nerror bars for the true mean.\n\nbootstrapped - Benefits\n-----------------------\n\n- Efficient computation of confidence intervals\n- Functions to handle single populations and a/b tests\n- Functions to understand `statistical\n power `__\n- Multithreaded support to speed-up bootstrap computations\n- Dense and sparse array support\n\nExample Usage\n-------------\n\n.. code:: python\n\n import numpy as np\n import bootstrapped.bootstrap as bs\n import bootstrapped.stats_functions as bs_stats\n\n mean = 100\n stdev = 10\n\n population = np.random.normal(loc=mean, scale=stdev, size=50000)\n\n # take 1k 'samples' from the larger population\n samples = population[:1000]\n\n print(bs.bootstrap(samples, stat_func=bs_stats.mean))\n >> 100.08 (99.46, 100.69)\n\n print(bs.bootstrap(samples, stat_func=bs_stats.std))\n >> 9.49 (9.92, 10.36)\n\nExtended Examples\n^^^^^^^^^^^^^^^^^\n\n- `Bootstrap\n Intro `__\n- `Bootstrap A/B\n Testing `__\n- More notebooks can be found in the\n `examples/ `__\n directory\n\nRequirements\n------------\n\n**bootstrapped** requires numpy. The power analysis functions require\nmatplotlib and pandas.\n\nInstallation\n------------\n\n.. code:: bash\n\n pip install bootstrapped\n\nHow bootstrapped works\n----------------------\n\n**bootstrapped** provides pivotal (aka empirical) based confidence\nintervals based on bootstrap re-sampling with replacement. The\npercentile method is also available.\n\nFor more information please see:\n\n1. `Bootstrap confidence\n intervals `__\n (good intro)\n2. `An introduction to Bootstrap\n Methods `__\n3. `The Bootstrap, Advanced Data\n Analysis `__\n4. `When the bootstrap dosen't\n work `__\n5. (book) `An Introduction to the\n Bootstrap `__\n6. (book) `Bootstrap Methods and their\n Application `__\n\nSee the CONTRIBUTING file for how to help out.\n\nContributors\n^^^^^^^^^^^^\n\nSpencer Beecher, Don van der Drift, David Martin, Lindsay Vass, Sergey\nGoder, Benedict Lim, and Matt Langner.\n\nSpecial thanks to Eytan Bakshy.\n\nLicense\n-------\n\n**bootstrapped** is BSD-licensed. We also provide an additional patent\ngrant.\n", "description_content_type": "", "docs_url": null, "download_url": "", "downloads": { "last_day": -1, "last_month": -1, "last_week": -1 }, "home_page": "https://github.com/facebookincubator/bootstrapped", "keywords": "", "license": "", "maintainer": "", "maintainer_email": "", "name": "bootstrapped", "package_url": "https://pypi.org/project/bootstrapped/", "platform": "", "project_url": "https://pypi.org/project/bootstrapped/", "project_urls": { "Homepage": "https://github.com/facebookincubator/bootstrapped" }, "release_url": "https://pypi.org/project/bootstrapped/0.0.2/", "requires_dist": null, "requires_python": "", "summary": "Implementations of the percentile based bootstrap", "version": "0.0.2" }, "last_serial": 4384673, "releases": { "0.0.1": [ { "comment_text": "", "digests": { "md5": "a6079487fb99062908c1d90bc64422b5", "sha256": "574bcb4978d26a5410b4ed13149ab7b41f6973d3b5a2cabee049d29261b5e841" }, "downloads": -1, "filename": "bootstrapped-0.0.1-py2.py3-none-any.whl", "has_sig": false, "md5_digest": "a6079487fb99062908c1d90bc64422b5", "packagetype": "bdist_wheel", "python_version": "3.5", "requires_python": null, "size": 12546, "upload_time": "2017-05-22T15:08:09", "url": "https://files.pythonhosted.org/packages/bf/5a/9a41920ee5403625216b9c3c77880534aa7c3afee45745d4abf0cabdacf2/bootstrapped-0.0.1-py2.py3-none-any.whl" }, { "comment_text": "", "digests": { "md5": "12989c50165fdcaa46ff2d4cb533d28b", "sha256": "ab27622e1badc63941c654075490ed47da94877d224f0f2d8d24d40ce6d324a3" }, "downloads": -1, "filename": "bootstrapped-0.0.1.tar.gz", "has_sig": false, "md5_digest": "12989c50165fdcaa46ff2d4cb533d28b", "packagetype": "sdist", "python_version": "source", "requires_python": null, "size": 7986, "upload_time": "2017-05-22T15:08:07", "url": "https://files.pythonhosted.org/packages/69/a3/de6395dfcffa8bbea78834117e87eedbeb955f17c5c770a9ea7ead27e020/bootstrapped-0.0.1.tar.gz" } ], "0.0.2": [ { "comment_text": "", "digests": { "md5": "9351d190a2a5b23cf893d2dc679df624", "sha256": "7d2cc81642c86dd3426518e1aa3e828f227ffb423fabaeaf4acc7e268d02f9e9" }, "downloads": -1, "filename": "bootstrapped-0.0.2.tar.gz", "has_sig": false, "md5_digest": "9351d190a2a5b23cf893d2dc679df624", "packagetype": "sdist", "python_version": "source", "requires_python": null, "size": 11588, "upload_time": "2018-10-17T02:23:10", "url": "https://files.pythonhosted.org/packages/b2/f7/beb22809fdc4781ae73d12c78e2cf14ae533ad9eec2fb84ec22e5b0b660c/bootstrapped-0.0.2.tar.gz" } ] }, "urls": [ { "comment_text": "", "digests": { "md5": "9351d190a2a5b23cf893d2dc679df624", "sha256": "7d2cc81642c86dd3426518e1aa3e828f227ffb423fabaeaf4acc7e268d02f9e9" }, "downloads": -1, "filename": "bootstrapped-0.0.2.tar.gz", "has_sig": false, "md5_digest": "9351d190a2a5b23cf893d2dc679df624", "packagetype": "sdist", "python_version": "source", "requires_python": null, "size": 11588, "upload_time": "2018-10-17T02:23:10", "url": "https://files.pythonhosted.org/packages/b2/f7/beb22809fdc4781ae73d12c78e2cf14ae533ad9eec2fb84ec22e5b0b660c/bootstrapped-0.0.2.tar.gz" } ] }