Data preparation

Enter your own data into the Data worksheet. The time series are organized by columns. The first column is always the time of observations, e.g., years. The first row must always contain labels (names of time series). These labels will be used to name new output worksheets. The time series in the data matrix can start and end at different years, except those intended for shift testing in the correlation.

There should not be any gaps (empty rows) within the time series, as in the case of missing data, because a gap indicates an end of the time series. If there is no way to fill the gaps, it is OK to delete empty rows. The first column is just data labels, and for the program to work, it is not necessary for the time steps between data points to be the same. For example, if the data starts in 1948 and the year of 1950 in missing, the data matrix will look like

When you run the test for shifts in mean, make sure the Data worksheet is active. For shifts in variance, run the test for shift in mean first, and then switch to Residuals worksheet making it active before you click the Run test button.

If you want to experiment with normally distributed random numbers, click Regime test --> Random Data (or Correlated Data) to populate the spreadsheet. Maximum number of time series is 250 for shifts in mean and variance, and 2 for shifts in the correlation.