| Title: | Change Point Analysis in ARIMA Forecasting |
| Version: | 0.20.0 |
| Author: | Claster William B. [aut], Philip Sallis [aut], Nhat Cuong Pham [aut, cre] |
| Maintainer: | Nhat Cuong Pham <acmetal74@gmail.com> |
| Description: | Package to incorporate change point analysis in ARIMA forecasting. |
| Depends: | R (≥ 3.1.2) |
| License: | GPL-3 |
| LazyData: | true |
| Suggests: | R.rsp |
| Imports: | changepoint, forecast, signal |
| VignetteBuilder: | R.rsp |
| RoxygenNote: | 5.0.1 |
| NeedsCompilation: | no |
| Packaged: | 2015-12-11 15:06:12 UTC; natto |
| Repository: | CRAN |
| Date/Publication: | 2016-09-16 12:50:19 |
CPI Function
Description
Incorporate change point analysis in ARIMA forecasting
Usage
cpi(myts, startChangePoint = 1, endChangePoint = 0, step = 1, num = 15,
cpmeth = "BinSeg", CPpenalty = "SIC", showModel = FALSE)
Arguments
myts |
a time series object |
startChangePoint |
a positive integer for minimum number of changepoints |
endChangePoint |
a positive integer for maximum number of change points. If 0 then only startChangePoint number of change points will be entered. Should be either 0 or greater than startChangePoint and if so the algorithm will loop through all values inbetween subject to step |
step |
an integer to step through loop of change points |
num |
Bump model number (see below) |
cpmeth |
changepoint method. Default is BinSeg. See cpa package for details |
CPpenalty |
default is SIC. See cpa package for details |
showModel |
default is False, if True shows all models for all changepoints, if an integer all models for that changepoint, if a string all changepoints for that model |
Value
A data frame with all the results from analysis