Time Variable Specs

On the Statistica Extract, Transform, and Load (ETL): ID-based dialog box -  Advanced  tab, click the Time Variable Specs button to display the Time Variable Specs dialog box, which provides options for the Time variable, if selected.

This dialog box provides options to:

    1. Normalize the lengths of batch data to achieve a fixed number of time intervals for all batches in the respective input data source.

    2. Optionally unstack the batch-time data. For each time interval and for each output variable, a new variable is created. This prepares data for Multivariate Statistical Process Control (MSPC).

Time interval definition. ID-based Statistica ETL aligns data differently from Time-indexed Statistica ETL, which aligns by minute, hour, day, week, month, quarter, or year.

ID-based Statistica ETL provides options to create identical numbers of elapsed time intervals for each batch (i.e. Class ID), when in the input data the number of such intervals is unequal for different batches.

    • None. This is the default option. Select this option to retain all date/time stamps. Batch sizes remain unchanged.

    • N-equal intervals. Select this option to divide every batch into a fixed number of equal (N) time intervals. You can specify 2 to 100 intervals in the text field provided. Intervals are computed using the method specified by the Computation of percentiles option, located on the Limits tab of the Options dialog box (accessed by clicking Options on the Tools tab or menu). The default method is Empirical Distribution Function w/Average.

    • N-user specified intervals. Select this option to divide every batch into N exact user-specified interval boundaries. You can specify 2 to 100 intervals in the text field provided. Click the Specify intervals button to display the User Specified Intervals dialog box, where you can enter values for the upper bounds.

Convert Date/Time Stamp to Elapsed Time. Select this check box to create a new time-based column with discrete interval values (i.e., 1, 2, 3, …). This column can be used by Multivariate Statistical Process Control (MSPC) to determine trends in the evolution of process variables (such as pressure, temperature, feed-rates, etc.)

Note: This check box is enabled when you have selected Date or Time as a Time Variable. However, when you have both Date and Time selected as Time Variables, this check box is disabled because elapsed time cannot be represented as date and time.

New time variable name. When the previous check box is selected, a new column is added with the default name "Time Step." The name can be changed in this edit box.

Unstack data by date/time. Select this check box to rotate (or pivot) time rows to time columns so that resultant data can be more easily manipulated by an analysis such as Multivariate Statistical Process Control (MSPC). For a detailed description of unstacking, see Unstacking/Stacking.

Replace missing values with previous values. Select this check box to replace any missing data with a duplicate of the previous observation. This check box is enabled when you select a Time interval definition option of N equal intervals or N user specified intervals.