Features

Time series

Model values that change over time, such as prices, demand, or rates, and simulate their paths.

Fit a time-series model to your data

Select a data series and click Time Series Fit. Probistica fits 13 models and ranks them by AIC or BIC. The forecast chart shows the history, the mean path, and the 25-75% and 5-95% bands.

  • Models: AR1, AR2, MA1, MA2, ARMA11, ARCH1, GARCH11, EGARCH11, APARCH11, GBM, GBM with jumps, BMMR, and BMMR with jumps.
  • Data steps: transform, detrend, and deseasonalize before the fit.
  • Start the forecast from the last value of the data.
  • Time Series Batch Fit fits many series at one time.
The Time Series Fit window. A table ranks AR1, BMMR, AR2, and other models by AIC, and a chart shows 60 historical values and a forecast fan of 12 periods.
Time Series Fit with the forecast of the best model.

Time-series functions in cells

Select a range of one column or one row, type the function, for example =SIM.AR1(100,40,0.8,490), and press Ctrl+Shift+Enter. Each cell is one period. Properties add transforms, integration, and seasonality. Time series can be correlated.

Limits of this release: the fit has no automatic detection, and some properties do not work on a series. The user guide lists them.