Type: Package
Title: Locating Distributional Changes in Highly Dependent Time Series
Version: 1.0.3
Maintainer: Lukas Zierahn <lukas@kappa-mm.de>
Description: Provides algorithms to locate multiple distributional change-points in piecewise stationary time series. The algorithms are provably consistent, even in the presence of long-range dependencies. Knowledge of the number of change-points is not required. The code is written in Go and interfaced with R.
License: GPL-2 | GPL-3 [expanded from: GPL]
URL: https://github.com/azalk/GoChest
BugReports: https://github.com/azalk/GoChest/issues
Imports: Rdpack, reticulate
Suggests: testthat
RdMacros: Rdpack
Encoding: UTF-8
LazyData: true
RoxygenNote: 7.1.1
NeedsCompilation: no
Packaged: 2021-02-13 15:46:50 UTC; lukas
Author: Lukas Zierahn [cre, aut], Azadeh Khaleghi [aut]
Repository: CRAN
Date/Publication: 2021-02-13 16:00:02 UTC

find_changepoints

Description

Returns the position of changepoints in the sequence. NOTE: PyChest needs to be installed first by calling ‘install_PyChest’.

Usage

find_changepoints(sample, minimum_distance, process_count)

Arguments

sample

A vector of floats corresponding to the piecewise stationary sample where the retrospective changes are to be sought

minimum_distance

A real number between 0 and 1 corresponding to a lower-bound on the minimum normalized length of the stationary segments (as percentage of total sample length)

process_count

The different number of distinct stationary processes present.

Value

The list of changepoints in increasing size

References

Khaleghi A, Ryabko D (2014). “Asymptotically consistent estimation of the number of change points in highly dependent time series.” In International Conference on Machine Learning, 539–547.

Khaleghi A, Ryabko D (2012). “Locating changes in highly dependent data with unknown number of change points.” In Advances in Neural Information Processing Systems, 3086–3094.


install_PyChest

Description

Initializes the package and installs/updates PyChest into the local reticulate-Python environment

Usage

install_PyChest()

Value

No return value, called to install the PyChest Package


list_estimator

Description

Returns the position of changepoints in the sequence. NOTE: PyChest needs to be installed first by calling ‘install_PyChest’.

Usage

list_estimator(sample, minimum_distance)

Arguments

sample

A vector of floats corresponding to the piecewise stationary sample where the retrospective changes are to be sought

minimum_distance

A real number between 0 and 1 corresponding to a lower-bound on the minimum normalized length of the stationary segments (as percentage of total sample length)

Value

The list of changepoints in order of score

References

Khaleghi A, Ryabko D (2012). “Locating changes in highly dependent data with unknown number of change points.” In Advances in Neural Information Processing Systems, 3086–3094.

mirror server hosted at Truenetwork, Russian Federation.