Anomalous time series package for R (ACM)
Switch branches/tags
Nothing to show
Clone or download
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
R Fixed trend calc for non-seasonal data Apr 17, 2018
data Removed empty columns from data May 26, 2015
demo Fixed demo Jun 2, 2015
inst/tests fix testthat dir problem May 27, 2015
man Added ordering option to ahull anomalies May 28, 2015
tests Corrected spelling of principal May 27, 2015
DESCRIPTION Changed name to conform to CRAN policies May 21, 2015
NAMESPACE anomalous pkg under ACM license May 12, 2015
README.md Updated readme Aug 26, 2017

README.md

Anomalous time-series R Package

It is becoming increasingly common for organizations to collect very large amounts of data over time, and to need to detect unusual or anomalous time series. For example, Yahoo has banks of mail servers that are monitored over time. Many measurements on server performance are collected every hour for each of thousands of servers. A common use-case is to identify servers that are behaving unusually. Methods in this package compute a vector of features on each time series, measuring characteristics of the series. For example, the features may include lag correlation, strength of seasonality, spectral entropy, etc. Then a robust principal component decomposition is used on the features, and various bivariate outlier detection methods are applied to the first two principal components. This enables the most unusual series, based on their feature vectors, to be identified. The bivariate outlier detection methods used are based on highest density regions and alpha-hulls. For demo purposes, this package contains both synthetic and real data from Yahoo.

A cut-down version of this package under a GPL licence is available from http://github.com/robjhyndman/anomalous.

Installation

You can install the package using

# install.packages("devtools")
devtools::install_github("robjhyndman/anomalous-acm")

Simple Example

  z <- ts(matrix(rnorm(3000),ncol=100),freq=4)
  y <- tsmeasures(z)
  biplot.features(y)
  anomaly(y)

License

This package is free and open source software, licensed under ACM.