R packages
Click on each item to see more details.

tsdecomp
Decomposition of Time Series Data

tsoutliers
Detection of Outliers in Time Series
This package implements a procedure based on the approach described
in Chen and Liu (1993,
doi:10.2307/2290724)
for automatic detection of outliers in time series.
Innovational outliers, additive outliers, level shifts, temporary changes
and seasonal level shifts are considered.
CRAN 
Document

bsm.airp
Fit the Basic Structural Model to the Airline Passengers Data
This package provides functions and scripts to replicate the results
discussed in the document 101 variations on a maximum likelihood procedure
to fit the basic structural time series model.
Source 
Reference guide 
Document

stsm.em
ExpectationMaximization Algorithm for Structural Time Series Models
The package implements the standard EM algorithm described in the litetature
for pure variance structural time series models. A modified version described in
the document linked below is also implemented. The computations of the
latter algorithm are suited to be computed in parallel.
A parallel implementation using Cilk plus is available in the package.
Source 
Document

temper
Functions for the Analysis of Temperature Trends in Alaska
This package contains the functions used to obtain the results discussed in
the paper "Trends in Alaska Temperature Data. Towards a More Realistic Approach",
Climate Dynamics.
Source 
DOI

stsm
Structural Time Series Models
This package provides some functions to fit time series models
by means of the traditional maximization of the time domain likelihood function
via quasiNewton algorithm. It is used as a reference to compare the
results based on the scoring algorithm to maximize the frequency domain
likelihood function.
CRAN 
Document 
code chunks

stsm.class
Class and Methods for Structural Time Series Models
This package defines a class and several methods to work with
structural time series models.
CRAN

KFKSDS
Kalman Filter, Kalman Smoother and Disturbance Smoother
This package provides a naive implementation of the Kalman filter,
Kalman smoother and disturbance smoother.
CRAN

meboot
Maximum Entropy Bootstrap for Time Series
The package implements the maximum entropy bootstrap proposed by Prof. H.D. Vinod.
CRAN 
JSS paper

partsm
Periodic Autoregressive Time Series Models
This package was developed as a pedagogical exercise while reading the book
Periodicity and Stochastic Trends in Economic Time Series
Franses, P.H.B.F. (1996)
(info).
The package replicates some of the results shown in the first eight chapters of the book.
Acknowledgements:
Many thanks to Matthieu Stigler for maintaining the package active on CRAN.
CRAN 
Document

uroot
Unit Root Tests for Seasonal Time Series
Seasonal unit roots and seasonal stability tests.
Pvalues based on response surface regressions are available for both tests.
Pvalues based on bootstrap are available for seasonal unit root tests.
A parallel implementation of the bootstrap method requires a CUDA capable GPU
with compute capability >= 3.0, otherwise a debugging version fully coded in R is used.
CRAN 
Document 1 
Document 2