Split Time Series
Time series can be divided into several time series according to various criteria using this method. On the one hand, it can be divided whenever a new day, week, month, or year begins, and on the other hand, when the time span between two data points meets a certain condition.
To split a time series, choose Methods → Cut/Extend → Further → Split from the ribbon (see figure above) or Methods → Change time axis → Split from the context menu of the time series in the Project Explorer (see the figure below) or the key combination <Ctrl>+<Alt>+<S>.
The dialog box of the method opens (see following figure).
Click Select time series and select from the open time series those to be split. A multiple selection is possible.
Select the Criterion for the division from the drop-down list: Split at specific intervals (see figure above) or Time difference between data points (see figure below).
Divide into Starting Days, Weeks, Months, or Years
To Split at specific intervals, select the interval length by clicking: Days, Weeks, Months, or Years. The following figure illustrates schematically the division of a time series into months.
The time series is divided as soon as a new interval begins.
Note that a high resolution can slow down the program due to the large number of time series to be created.
Dividing a time series of one year into Months produces 12 time series (one time series per month).
The division into Weeks generates 52 time series (one time series per week).
Split at Specific Time Steps
Dividing according to the split condition Time difference between data points occurs at all points in the time series at which the specified time difference between two data points (step size) exists.
You can freely select the Time difference using the operators greater or equal, equal, and smaller or equal (see following figure).
The prerequisite is that the selected interval actually exists in the time series, otherwise no position for separation can be found.
An unequally spaced time series is divided into steps of hours except at one point. At this point, the interval between two data points is two hours.
This time series is divided by the Time difference greater than or equal to two hours.
Two new time series are created. The division occurs at the point where the time difference between two data points is two hours.
For equally spaced time series, the selected time difference can only be identical with the step size of the time series. Therefore, the division of an equally spaced time series results in as many time series as the time series contains time steps.
With the Split method, the original time series cannot be overwritten, but new result time series are always generated. A Suffix can be defined for the names of the new time series to be created.
