%global _empty_manifest_terminate_build 0 Name: python-odds Version: 0.2.2 Release: 1 Summary: outlier_detection, detects outliers. License: MIT License URL: https://github.com/jogrundy/od_dist Source0: https://mirrors.aliyun.com/pypi/web/packages/4a/b9/70317d44d77a2865bf8ed79d73cadb3b2d0661f23e4887ebb1a555203114/odds-0.2.2.tar.gz BuildArch: noarch Requires: python3-numpy Requires: python3-scipy Requires: python3-scikit-learn Requires: python3-torch %description [![Build Status](https://app.travis-ci.com/jogrundy/od_dist.svg?branch=main)](https://app.travis-ci.com/jogrundy/od_dist) # Outlier Data Detection Systems - ODDS # As used in paper "Simple Models are Effective in Anomaly Detection in Multi-variate Time Series" This work was funded by the grant "Early detection of contact distress for enhanced performance monitoring and predictive inspection of machines" (EP/S005463/1) from the Engineering and Physical Sciences Research Council (EPSRC), UK. > pip install odds The work is done by the **OD** object. Import the 'OD' object as follows: > from odds import OD Instantiate the object with the 'algo' argument, where a short string represents the algorithm you wish to use. In this case, 'VAR' refers to vector autoregression, a simple linear multidimensional regression algorithm. Other implemented algorithms are listed below. > od = OD('VAR') To use the object, you need to call the 'get_os()' function, with 'X' as its argument, where X is a data matrix, **n** samples by **p** features. **p must be 2 or greater to work** on many of the systems, this returns a vector with n scores, one for each sample. > outlier_scores = od.get_os(X) The higher scores are the more outlying. you can then set a threshold if you wish, or just look at the ranking. Scores have not been sanitised, they may contain 'nan' values particularly from the 'VAE' if the data input has not been scaled. However it seems other algorithms work better without scaling, so inputs are not automatically scaled. Hyperparameters for each of these algorithms are currently fixed to the values in my paper, however at some point I will be finishing implementing a pass though allowing you to specify the hyperparameters at instantiation. This is on my ToDo list. To get normalised (between 0 and 1) scores, use the 'norm' keyword argument. This may result in errors if the data is not normalised, as there may be infinite values in the scores (usually only from the 'VAE'). > normalised_scores = od.get_os(X, norm=True) Valid strings for outlier algorithms: - 'VAR' Vector Autoregression - 'FRO' Ordinary Feature Regression - 'FRL' LASSO Feature Regression - 'FRR' Ridge Feature Regression - 'GMM' Gaussian Mixture model - 'IF' Isolation Forest - 'DBSCAN' Density Based Spatial Clustering and Noise - 'OCSVM' One Class Support Vector Machine - 'LSTM' Long Short Term Memory - 'GRU' Gated Recurrent Unit - 'AE' Autoencoder - 'VAE' Variational Autoencoder - 'OP' Outlier Pursuit - 'GOP' Graph Regularised Outlier Pursuit - 'RAND' Random scoring (for baseline comparison) Hyperparameter table ![Hyperparameter table](images/table_hyperparameters.png) %package -n python3-odds Summary: outlier_detection, detects outliers. Provides: python-odds BuildRequires: python3-devel BuildRequires: python3-setuptools BuildRequires: python3-pip %description -n python3-odds [![Build Status](https://app.travis-ci.com/jogrundy/od_dist.svg?branch=main)](https://app.travis-ci.com/jogrundy/od_dist) # Outlier Data Detection Systems - ODDS # As used in paper "Simple Models are Effective in Anomaly Detection in Multi-variate Time Series" This work was funded by the grant "Early detection of contact distress for enhanced performance monitoring and predictive inspection of machines" (EP/S005463/1) from the Engineering and Physical Sciences Research Council (EPSRC), UK. > pip install odds The work is done by the **OD** object. Import the 'OD' object as follows: > from odds import OD Instantiate the object with the 'algo' argument, where a short string represents the algorithm you wish to use. In this case, 'VAR' refers to vector autoregression, a simple linear multidimensional regression algorithm. Other implemented algorithms are listed below. > od = OD('VAR') To use the object, you need to call the 'get_os()' function, with 'X' as its argument, where X is a data matrix, **n** samples by **p** features. **p must be 2 or greater to work** on many of the systems, this returns a vector with n scores, one for each sample. > outlier_scores = od.get_os(X) The higher scores are the more outlying. you can then set a threshold if you wish, or just look at the ranking. Scores have not been sanitised, they may contain 'nan' values particularly from the 'VAE' if the data input has not been scaled. However it seems other algorithms work better without scaling, so inputs are not automatically scaled. Hyperparameters for each of these algorithms are currently fixed to the values in my paper, however at some point I will be finishing implementing a pass though allowing you to specify the hyperparameters at instantiation. This is on my ToDo list. To get normalised (between 0 and 1) scores, use the 'norm' keyword argument. This may result in errors if the data is not normalised, as there may be infinite values in the scores (usually only from the 'VAE'). > normalised_scores = od.get_os(X, norm=True) Valid strings for outlier algorithms: - 'VAR' Vector Autoregression - 'FRO' Ordinary Feature Regression - 'FRL' LASSO Feature Regression - 'FRR' Ridge Feature Regression - 'GMM' Gaussian Mixture model - 'IF' Isolation Forest - 'DBSCAN' Density Based Spatial Clustering and Noise - 'OCSVM' One Class Support Vector Machine - 'LSTM' Long Short Term Memory - 'GRU' Gated Recurrent Unit - 'AE' Autoencoder - 'VAE' Variational Autoencoder - 'OP' Outlier Pursuit - 'GOP' Graph Regularised Outlier Pursuit - 'RAND' Random scoring (for baseline comparison) Hyperparameter table ![Hyperparameter table](images/table_hyperparameters.png) %package help Summary: Development documents and examples for odds Provides: python3-odds-doc %description help [![Build Status](https://app.travis-ci.com/jogrundy/od_dist.svg?branch=main)](https://app.travis-ci.com/jogrundy/od_dist) # Outlier Data Detection Systems - ODDS # As used in paper "Simple Models are Effective in Anomaly Detection in Multi-variate Time Series" This work was funded by the grant "Early detection of contact distress for enhanced performance monitoring and predictive inspection of machines" (EP/S005463/1) from the Engineering and Physical Sciences Research Council (EPSRC), UK. > pip install odds The work is done by the **OD** object. Import the 'OD' object as follows: > from odds import OD Instantiate the object with the 'algo' argument, where a short string represents the algorithm you wish to use. In this case, 'VAR' refers to vector autoregression, a simple linear multidimensional regression algorithm. Other implemented algorithms are listed below. > od = OD('VAR') To use the object, you need to call the 'get_os()' function, with 'X' as its argument, where X is a data matrix, **n** samples by **p** features. **p must be 2 or greater to work** on many of the systems, this returns a vector with n scores, one for each sample. > outlier_scores = od.get_os(X) The higher scores are the more outlying. you can then set a threshold if you wish, or just look at the ranking. Scores have not been sanitised, they may contain 'nan' values particularly from the 'VAE' if the data input has not been scaled. However it seems other algorithms work better without scaling, so inputs are not automatically scaled. Hyperparameters for each of these algorithms are currently fixed to the values in my paper, however at some point I will be finishing implementing a pass though allowing you to specify the hyperparameters at instantiation. This is on my ToDo list. To get normalised (between 0 and 1) scores, use the 'norm' keyword argument. This may result in errors if the data is not normalised, as there may be infinite values in the scores (usually only from the 'VAE'). > normalised_scores = od.get_os(X, norm=True) Valid strings for outlier algorithms: - 'VAR' Vector Autoregression - 'FRO' Ordinary Feature Regression - 'FRL' LASSO Feature Regression - 'FRR' Ridge Feature Regression - 'GMM' Gaussian Mixture model - 'IF' Isolation Forest - 'DBSCAN' Density Based Spatial Clustering and Noise - 'OCSVM' One Class Support Vector Machine - 'LSTM' Long Short Term Memory - 'GRU' Gated Recurrent Unit - 'AE' Autoencoder - 'VAE' Variational Autoencoder - 'OP' Outlier Pursuit - 'GOP' Graph Regularised Outlier Pursuit - 'RAND' Random scoring (for baseline comparison) Hyperparameter table ![Hyperparameter table](images/table_hyperparameters.png) %prep %autosetup -n odds-0.2.2 %build %py3_build %install %py3_install install -d -m755 %{buildroot}/%{_pkgdocdir} if [ -d doc ]; then cp -arf doc %{buildroot}/%{_pkgdocdir}; fi if [ -d docs ]; then cp -arf docs %{buildroot}/%{_pkgdocdir}; fi if [ -d example ]; then cp -arf example %{buildroot}/%{_pkgdocdir}; fi if [ -d examples ]; then cp -arf examples %{buildroot}/%{_pkgdocdir}; fi pushd %{buildroot} if [ -d usr/lib ]; then find usr/lib -type f -printf "\"/%h/%f\"\n" >> filelist.lst fi if [ -d usr/lib64 ]; then find usr/lib64 -type f -printf "\"/%h/%f\"\n" >> filelist.lst fi if [ -d usr/bin ]; then find usr/bin -type f -printf "\"/%h/%f\"\n" >> filelist.lst fi if [ -d usr/sbin ]; then find usr/sbin -type f -printf "\"/%h/%f\"\n" >> filelist.lst fi touch doclist.lst if [ -d usr/share/man ]; then find usr/share/man -type f -printf "\"/%h/%f.gz\"\n" >> doclist.lst fi popd mv %{buildroot}/filelist.lst . mv %{buildroot}/doclist.lst . %files -n python3-odds -f filelist.lst %dir %{python3_sitelib}/* %files help -f doclist.lst %{_docdir}/* %changelog * Thu Jun 08 2023 Python_Bot - 0.2.2-1 - Package Spec generated