Missing data treatment substitutes empty series values using a variety of missing data substitution methods. Methods are included in the group of methods that perform transformations on series.
The following methods are available in the time series analysis tool:
Geometric Interpolation. Empty values of the series are calculated using geometric interpolation.
Cubic Spline Interpolation. Empty values of the series are calculated using cubic spline interpolation.
Linear Interpolation. Empty values of series are calculated using linear interpolation by two neighboring points.
Linear Trend. Empty values of series are calculated using linear trend.
Previous Value. Empty values are replaced with the previous non-empty value.
Succeeding Value. Empty series values are substituted with the next non-empty value.
Average. Empty values of the series are calculated as average.
N-Point Average. Empty values of series are calculated as N-point average. The number of N points can be specified on the Parameters side panel tab.
Previous Growth Rate. Empty values of series are calculated based on the growth rate of existing values in comparison with the previous period. The number of periods can be specified on the Parameters side panel tab.
Succeeding Growth Rate. Empty series values are calculated based on the growth rate of existing values in comparison with the succeeding period. The number of periods can be specified on the Parameters side panel tab.
Value. The number specified on the Parameters side panel tab substitutes empty values of series.
NOTE. The Linear Trend, Average, N-Point Average methods are available only in the desktop application.
After the method is applied, in the workbook, based on each selected series a calculated series with a name of the <Method_Name>(<Series_Name>) type is created, containing the calculation results. For example:
To set up parameters of the method of missing data substitution, use the Parameters side panel tab.
Set the parameter value. The calculated series is recalculated.
See also:
Working with Calculated Series | Methods of Missing Data Treatment | Modeling Container Missing Data Treatment | IModelling.Fill