The GROUPING operator enables data grouping operations on the input data bus stream. The operator works by extracting unique combinations of values in selected columns (bits), which together form the so-called grouping key.
For each unique sequence of values in the grouping key columns, a single output record is generated. Columns that are not part of the grouping key must be processed using a specified aggregation function, such as:
- maximum value,
- minimum value,
- sum,
- average,
- count of occurrences.
The operator processes the entire dataset available at the input, enabling data consolidation in aggregated form.
Typical use cases of the GROUPING operator include:
- removing duplicates based on selected criteria,
- generating summary reports,
- preparing data for further processing in a simplified form.
An example of a GRAVITY project using the GROUPING operator:
GROUPING RULE CONFIGURATION
You can define a grouping key consisting of any number of data bus columns using a drag-and-drop action (you drag a column from the input data bus and drop it onto the grouping key list). Additionally, for the remaining columns of the input data bus, you define an aggregation function using the BEHAVIOUR TYPE setting.
MAXMINSUMCOUNTAVG
At the operator’s output, you will have a processed data stream with the same width (that is, the same number of columns) as the set of grouped and aggregated columns.