Metadata-Version: 2.1
Name: pybats-detection
Version: 0.2.1
Summary: Tools for outlier and structural changes detection in time series analysis using Bayesian Dynamic Linear Model.
Home-page: https://github.com/Murabei-OpenSource-Codes/pybats-detection
Author: André Menezes and Eduardo Pinheiro
Author-email: andrefelipemaringa@gmail.com
License: UNKNOWN
Description: ## Welcome to `pybats-detection`
        
        The `pybats-detection` is a `python` package with routines implemented in python for detection of outlier and structural changes in time series using Bayesian Dynamic Linear Models (DLM).
        The currently version of the package implements the automatic monitoring, manual intervention and smoothing for DLM’s.
        
        The stable version of `pybats-detection` can be installed from [PyPI](https://pypi.org/) using:
        
        ```
        pip install pybats-detection
        ```
        
        The development version can be installed from [GitHub](https://github.com/) using:
        
        ```
        git clone git@github.com:Murabei-OpenSource-Codes/develop/pybats-detection.git pybats-detection
        cd pybats-detection
        python setup.py install
        ```
        
        The package uses the `pybats.dglm.dlm` objects from [`PyBATS`](https://github.com/lavinei/pybats) package as an input for the following classes:
        
        - `Monitoring`: perform automatic monitoring of outlier and/or structural changes in time series according to [West and Harisson (1986)](https://www.tandfonline.com/doi/abs/10.1080/01621459.1986.10478331) .
        
        - `Intervention`: perform manual intervention of outlier and/or structural changes in time series according to [West and Harrison (1989)](https://onlinelibrary.wiley.com/doi/abs/10.1002/for.3980080104).
        
        - `Smoothing`: compute the retrospective state space parameter and predictive distributions.
        
        All three classes have the `fit` method which received the univariate time series
        as a `pandas.Series` object and further arguments related to each class.
        
        User manuals can be found in:
        
        - [pybats_detection](https://raw.githubusercontent.com/Murabei-OpenSource-Codes/pybats-detection/develop/vignettes/pybats_detection.pdf): detailed explanation of `pybats-detection` usability.
        
        - [quick_start](https://raw.githubusercontent.com/Murabei-OpenSource-Codes/pybats-detection/develop/vignettes/quick_start.pdf): quick reference guide with step-by-step usability.
        
        ## Authors
        
        `pybats-detection` was developed by [André Menezes](https://andrmenezes.github.io/) and
        [Eduardo Gabriel](https://www.linkedin.com/in/eduardo-gabriel-433332142/)
        while working as Data Scientist at [Murabei Data Science](https://www.murabei.com/)
        advised by professor [Hélio Migon](http://lattes.cnpq.br/7997248190492823) and
        [André Baceti](https://br.linkedin.com/in/andre-baceti/pt) .
        
        
        ## License
        
        The `pybats-detection` package is released under the Apache License, Version 2.0.
        Please, see file [`LICENSE.md`](https://github.com/Murabei-OpenSource-Codes/pybats-detection/blob/develop/LICENSE.md).
        
Platform: UNKNOWN
Classifier: Programming Language :: Python :: 3
Classifier: License :: OSI Approved :: Apache Software License
Classifier: Operating System :: OS Independent
Requires-Python: >=3.6
Description-Content-Type: text/markdown
