Time Series and Forecasting

  • Time Series Plots
    Creates plots of time series; lists of variables can be plotted in a single graph, against a single axis or against two differently scaled axes, or they can be plotted in multiple graphs, one variable at a time.
  • Single-Series Transformations (x=f(x))
    Creates single-series transformations of the kind x=f(x) for selected continuous variables; computes a summary graph and spreadsheet of the series after performing transformation; autocorrelation functions can also be computed. Available transformations include: Adding a constant, power transformation, inverse power transformation, natural log transformation, exponent transformation, mean subtraction, standardization, trend subtraction, and autocorrelation subtraction.
  • Two-Series Transformation (x=f(x,y))
    Creates two-series transformations of the kind x=f(x,y) for selected continuous variables; computes a summary graph and spreadsheet of the series after computing the transformation; autocorrelation functions can also be computed. Available transformations include: Differencing (x=x-y) with or without lag Residualizing (x=x-(a+b*y(lag))).
  • Differencing, Time Series Transformations
    Differences the selected continuous variables, by a user-defined lag; computes a summary graph and spreadsheet of the series after differencing; autocorrelation functions can also be computed.
  • Smoothing Transformations
    Creates various smoothing transformations for time series data. Available methods include moving average and moving median smoothing, simple exponential smoothing (see also the designated seasonal and nonseasonal exponential smoothing facilities), and the 4253H filter.
  • Simple Fourier-Type Transformations
    Creates various transformations related to Fourier (spectral) decomposition, including tapering, various smoothing methods (Daniell, Tukey, Hamming, Parzen, Bartlett), the real and imaginary parts of the series, as well as the inverse transformation. See also the Single Series Fourier Analysis and Two Series Fourier Analysis facilities for a complete set of options for spectral analysis.
  • Autocorrelations and Crosscorrelations
    Creates and plots the autocorrelation and partial autocorrelation functions; if crosscorrelations are requested, the crosscorrelation function for all pairs of continuous dependent variables is also computed. Results include the autocorrelations, their standard errors, the so-called Box-Ljung statistic, and the significance level of that statistic.
  • Distributed Lags Analysis
    Performs a complete distributed lags analysis for all variables in the continuous dependent variable list, treating the (single) continuous predictor variable as the independent variable. Various summary reports and plots are available.
  • Exponential Smoothing
    Performs simple and complex (multiple-parameter) exponential smoothing; models can include additive and multiplicative seasonal components, and linear, exponential, and damped trends components.
  • ARIMA Models
    Fits seasonal and non-seasonal ARIMA (p, d, q)(pS, dS, qS) models to continuous variables; various results graphs, forecasts, and tools for assessing the quality fit to the data are reported by the program.
  • Interrupted ARIMA
    Fits seasonal and non-seasonal ARIMA (p, d, q)(pS, dS, qS) models to continuous variables, and estimates parameters for one or more interventions (discontinuities) of different types; various results graphs, forecasts, and tools for assessing the quality fit to the data are reported by the program.
  • Single Series Spectral (Fourier) Analysis
    Performs a single series spectrum (Fourier) analysis for the specified variables (series); computes the sine and cosine coefficients, periodogram, and (smoothed) spectral density estimates using various types of weights (Daniell, Tukey, Hamming, Parzen, Bartlett).
  • Two Series Spectral (Fourier) Analysis
    Creates two-series cross-spectrum analyses, and reports the standard periodogram, density, cross-spectrum, cross-density, and other estimates (gain, cross-amplitude, etc.)
  • Seasonal Decomposition (Census I)
    Performs classical seasonal decomposition (Census method I; ratio-to-moving averages method); computes moving averages, ratios or differences, seasonal factors, seasonally adjusted series, the smoothed trend-cycle component, and the irregular component.
  • X11/Y2K Census Method II Monthly
    Provides a full-featured implementation of the US Bureau of the Census X-11 variant of the Census method II seasonal adjustment procedure. The arrangement of options follows closely the definitions and conventions described in the bureau of the Census documentation. Note that, unlike the Bureau of the Census implementation of this algorithm, the implementation in Statistica can handle dates past the year 2000.
  • X11/Y2K Census Method II Quarterly
    Provides a full-featured implementation of the US Bureau of the Census X-11 variant of the Census method II quarterly adjustment procedure. The arrangement of options follows closely the definitions and conventions described in the bureau of the Census documentation. Note that, unlike the Bureau of the Census implementation of this algorithm, the implementation in Statistica can handle dates past the year 2000.