Function reference • fastcpd

Main function

All implementation of fastcpd is unified into one single function fastcpd().

fastcpd(): Find change points efficiently

Wrapper functions

Simplified functions are provided for each family.

Time series

fastcpd.ts() and fastcpd_ts() are the main functions for time series data.

fastcpd.ar() fastcpd_ar(): Find change points efficiently in AR(p) models

fastcpd.arima() fastcpd_arima(): Find change points efficiently in ARIMA(p, d, q) models

fastcpd.arma() fastcpd_arma(): Find change points efficiently in ARMA(p, q) models

fastcpd.garch() fastcpd_garch(): Find change points efficiently in GARCH(p, q) models

fastcpd.ma() fastcpd_ma(): Find change points efficiently in MA(q) models

fastcpd.ts() fastcpd_ts(): Find change points efficiently in time series data

fastcpd.var() fastcpd_var(): Find change points efficiently in VAR(p) models

Unlabeled data

Used for data without response variables, for example, mean change and variance change.

fastcpd.mean() fastcpd_mean(): Find change points efficiently in mean change models

fastcpd.variance() fastcpd_variance(): Find change points efficiently in variance change models

fastcpd.meanvariance() fastcpd_meanvariance() fastcpd.mv() fastcpd_mv(): Find change points efficiently in variance change models

Regression data

Detect change points in the coefficients of regression-type data.

fastcpd.binomial() fastcpd_binomial(): Find change points efficiently in logistic regression models

fastcpd.lasso() fastcpd_lasso(): Find change points efficiently in penalized linear regression models

fastcpd.lm() fastcpd_lm(): Find change points efficiently in linear regression models

fastcpd.poisson() fastcpd_poisson(): Find change points efficiently in Poisson regression models

Utility functions

The following functions help with visualization and analyzation of the data.

plot(<fastcpd>) plot(<fastcpd>,<missing>): Plot the data and the change points for a fastcpd object

print(<fastcpd>) print(<fastcpd>): Print the call and the change points for a fastcpd object

show(<fastcpd>) show(<fastcpd>): Show the available methods for a fastcpd object

summary(<fastcpd>) summary(<fastcpd>): Show the summary of a fastcpd object

Data

fastcpd comes with a selection of built-in datasets that are used in examples to illustrate various change point detection challenges.

bitcoin: Bitcoin Market Price (USD)

occupancy: Occupancy Detection Data Set

transcriptome: Transcription Profiling of 57 Human Bladder Carcinoma Samples

uk_seatbelts: UK Seatbelts Data

well_log: Well-log Dataset from Numerical Bayesian Methods Applied to Signal Processing

Main class

The results of fastcpd() will be stored in the class fastcpd with accompanied several utility functions.

fastcpd-class: An S4 class to store the output created with fastcpd