Time series

GenStat provides several methods for examining and analysing time series. Sample correlation functions are produced by the directive CORRELATE:

forms correlations between variates, autocorrelations of variates, and lagged cross-correlations between variates

The analysis of Box-Jenkins models is specified by several directives:

forms preliminary estimates of parameters in time-series models

specifies input series and transfer-function models for subsequent estimation of a model for an output series

estimates parameters in Box-Jenkins models for time series

Information can be saved in GenStat data structures, or further output can be produced:

displays further output after an analysis by ESTIMATE

saves results after an analysis by ESTIMATE

forecasts future values of a time series

displays characteristics of a time series model

It is also possible to filter a time series, or perform spectral analysis via the Fourier transform of a time series using the directives:

filters time series by time-series models

calculates cosine or Fourier transforms of a real or complex series

Procedures in module timeseries of the Library include:

fits an ARIMA model, with forecasts and residual checks

plots forecasts of a time series using a previously fitted ARIMA

displays time series statistics useful for ARIMA model selection

gives periodogram-based tests for white noise in time series

filters a time series before spectral analysis

gives periodogram-based analyses for replicated time series

forms smoothed spectrum estimates for univariate time series