The forecast package for R provides methods and tools for displaying and analysing univariate time series forecasts including exponential smoothing via state space models and automatic ARIMA modelling. It also includes a handful of data sets from the Time Series Data Library. The package is described in Hyndman and Khandakar (2008).
The latest (beta) version is on github which is normally slightly ahead of the CRAN version (below).
Please report bugs or feature requests on github