Name | likelycause2 JSON |

Version | 0.1.8 JSON |

download | |

home_page | https://github.com/Ana-c-ng/likelycause2 |

Summary | Likely cause finds creative ways to identify causes |

upload_time | 2020-10-23 17:30:34 |

maintainer | |

docs_url | None |

author | Ana Garcia |

requires_python | >=2.7 |

license | MIT |

keywords | |

VCS | |

bugtrack_url | |

requirements | No requirements were recorded. |

Travis-CI | No Travis. |

coveralls test coverage | No coveralls. |

# Likelycause2 Likelycause is an utility package that uses several functions to attribute causes to variations. Using a combination of arithmetical decompositions and bayesian techniques, this was built to facilitate the workflow of a data-analyst working for the private sector. ## What the package contains This package has everything built under the likelycause2 module, so all the functions should be called using “likelycause2.”. Currently, we have 1 auxiliary function and 1 causal function. ### Auxiliary functions - likelycause2.last_period: The last period function is a utility function that builds variation variables in a dataframe._ ### Causal functions - likelycause2.bayes_suspects: The bayes_suspects function calculates the conditional probabilities of the event and each suspicious causes or a combination of those causes. It also suggests an attribution to each individual cause, by adjusting the intersections of causes ## Likelycause2.last_period ### Description: The last period function is a utility function that builds variation variables in a dataframe. Variations are defined between moment t and a moment in the past. ### Arguments: - df (pd.DataFrame): the dataframe - unique_id (string): unique identifier of each line. Must be unique, and can only be 1 column - interval (string): what is the interval you want to calculate variations for. Accepts days, weeks and hours - periods (int): number of periods you want to look back on that interval. For last variations, for example, the argument period would be 1 - date_column (string): the date column in your dateframe. Must be a datetime. To convert, use pandas.to_datetime function - to_past (list): list of columns you want to calculate the variations for ### Returns: Returns the dataframe that was inputed with additinal columns named v+name of the columns in the to_past argument. Those columns represent the variation of that variable between moment t and t-periods. This variation is calculated as (Variable in moment t)/(Variable in moment t-periods). ## Likelycause2.bayes_suspects ### Description: The bayes_suspects function calculates the conditional probabilities of the event and each suspicious causes or a combination of those causes. It also suggests an attribution to each individual cause, by adjusting the intersections of causes ### Arguments: - df (pd.DataFrame): the dataframe - event (string): name of the column that contains the event that we want to explain - suspects (list): list with name of the columns that contains the potential causes for what we want to explain - point (dictionary): dictionary with the point for which we want to calculate the probability. Must be a combination of the cause and all the individual points of suspects ### Returns: Returns a dataframe with all the possible probabilities combinations, and the conditional probabilities: - name: name of that conditional combination. If it has one event, it represents P(event|a). If it has 2 events it represents P(event1 & event2|a) - prob_ba: P(cause | event) - prob_a: P(cause) - prob_b: P(event) - pbayes: confitional probability - pbayes_attribution: suggested probability attribution if we want to attribute to individual causes

{ "_id": null, "home_page": "https://github.com/Ana-c-ng/likelycause2", "name": "likelycause2", "maintainer": "", "docs_url": null, "requires_python": ">=2.7", "maintainer_email": "", "keywords": "", "author": "Ana Garcia", "author_email": "ana.n.garcia2@gmail.com", "download_url": "https://files.pythonhosted.org/packages/66/2d/bddfedbbbbb03c0e38c5713944b0bd63a9afe8924fe170806019f014d4e8/likelycause2-0.1.8.tar.gz", "platform": "", "description": "# Likelycause2\n\nLikelycause is an utility package that uses several functions to attribute causes to variations. Using a combination of arithmetical decompositions and bayesian techniques, this was built to facilitate the workflow of a data-analyst working for the private sector.\n\n## What the package contains\nThis package has everything built under the likelycause2 module, so all the functions should be called using \u201clikelycause2.\u201d. Currently, we have 1 auxiliary function and 1 causal function.\n\n### Auxiliary functions\n- likelycause2.last_period: The last period function is a utility function that builds variation variables in a dataframe._\n\n### Causal functions\n- likelycause2.bayes_suspects: The bayes_suspects function calculates the conditional probabilities of the event and each suspicious causes or a combination of those causes. It also suggests an attribution to each individual cause, by adjusting the intersections of causes\n\n## Likelycause2.last_period\n\n### Description:\nThe last period function is a utility function that builds variation variables in a dataframe.\nVariations are defined between moment t and a moment in the past.\n\n### Arguments:\n\n- df (pd.DataFrame): the dataframe\n- unique_id (string): unique identifier of each line. Must be unique, and can only be 1 column\n- interval (string): what is the interval you want to calculate variations for. Accepts days, weeks and hours\n- periods (int): number of periods you want to look back on that interval. For last variations, for example, the argument period would be 1\n- date_column (string): the date column in your dateframe. Must be a datetime. To convert, use pandas.to_datetime function\n- to_past (list): list of columns you want to calculate the variations for\n\n### Returns:\nReturns the dataframe that was inputed with additinal columns named v+name of the columns in the to_past argument. Those columns represent the variation of that variable between moment t and t-periods. This variation is calculated as (Variable in moment t)/(Variable in moment t-periods).\n\n\n## Likelycause2.bayes_suspects\n\n### Description:\nThe bayes_suspects function calculates the conditional probabilities of the event and each suspicious causes or a combination of those causes. \nIt also suggests an attribution to each individual cause, by adjusting the intersections of causes\n\n### Arguments:\n\n- df (pd.DataFrame): the dataframe\n- event (string): name of the column that contains the event that we want to explain\n- suspects (list): list with name of the columns that contains the potential causes for what we want to explain\n- point (dictionary): dictionary with the point for which we want to calculate the probability. Must be a combination of the cause and all the individual points of suspects\n\n### Returns:\nReturns a dataframe with all the possible probabilities combinations, and the conditional probabilities:\n\n- name: name of that conditional combination. If it has one event, it represents P(event|a). If it has 2 events it represents P(event1 & event2|a)\n- prob_ba: P(cause | event)\n- prob_a: P(cause)\n- prob_b: P(event)\n- pbayes: confitional probability\n- pbayes_attribution: suggested probability attribution if we want to attribute to individual causes\n\n\n", "bugtrack_url": null, "license": "MIT", "summary": "Likely cause finds creative ways to identify causes", "version": "0.1.8", "split_keywords": [], "urls": [ { "comment_text": "", "digests": { "md5": "24ee1e105a936af6f0809d9891e39d03", "sha256": "04e7b035883c8fd1c8d1844790a5a83694b7f68cb492ccb9e23ff29f3ee1b67c" }, "downloads": -1, "filename": "likelycause2-0.1.8-py3-none-any.whl", "has_sig": false, "md5_digest": "24ee1e105a936af6f0809d9891e39d03", "packagetype": "bdist_wheel", "python_version": "py3", "requires_python": ">=2.7", "size": 5768, "upload_time": "2020-10-23T17:30:32", "upload_time_iso_8601": "2020-10-23T17:30:32.593258Z", "url": "https://files.pythonhosted.org/packages/c7/92/eb81054c7c2865a5ee782f7c2e5cf4365dd256d2f9ba43a492cfe3c07ab1/likelycause2-0.1.8-py3-none-any.whl", "yanked": false, "yanked_reason": null }, { "comment_text": "", "digests": { "md5": "66068ecd9aa183db14f741a8158fff8a", "sha256": "c4c848833f6382f8a46f7098edd660d02700e3f10069ae9cdf0e4dfc6611b6ca" }, "downloads": -1, "filename": "likelycause2-0.1.8.tar.gz", "has_sig": false, "md5_digest": "66068ecd9aa183db14f741a8158fff8a", "packagetype": "sdist", "python_version": "source", "requires_python": ">=2.7", "size": 4329, "upload_time": "2020-10-23T17:30:34", "upload_time_iso_8601": "2020-10-23T17:30:34.098366Z", "url": "https://files.pythonhosted.org/packages/66/2d/bddfedbbbbb03c0e38c5713944b0bd63a9afe8924fe170806019f014d4e8/likelycause2-0.1.8.tar.gz", "yanked": false, "yanked_reason": null } ], "upload_time": "2020-10-23 17:30:34", "github": true, "gitlab": false, "bitbucket": false, "github_user": null, "github_project": "Ana-c-ng", "error": "Could not fetch GitHub repository", "lcname": "likelycause2" }

Elapsed time: 0.15542s