‘pcts’ is an R package for modelling periodically correlated and periodically integrated time series.
Install the latest
stable version of pcts from CRAN:
install.packages("pcts")You can install the development version of
pcts from Github:
library(remotes)
install_github("GeoBosh/pcts")Periodic time series can be created with pcts(). Models
are fitted with fitPM() and several other functions. To
obtain periodic properties, such as sample periodic autocorrelations of
periodic time series or theoretical periodic autocorrelations of
periodic models, just call the respective functions (here
autocorrelations() and
partialAutocorrelations()) and they will compute the
relevant property depending on the class of the argument, see the
examples in the documentation.
A good place to start is the help topic ?pcts-package.
Several datasets are available for examples and experiments. For
example, ?dataFranses1996 contains the data from Franses
(1996). The datasets are from classes "mts" or
"ts" (the standard R classes for time series), so can be
used without loading pcts, if desired.