Which statement best describes the relationship between pre-sorted data and performance when using Process When Group Changes?

• A. Using Process When Group Changes with unsorted data yields best performance.
• B. Pre-sorting data before using Process When Group Changes is recommended for performance.
• C. Blocking mode provides the best performance with pre-sorted data.
• D. Group By has no impact on performance.

Answer: (B)

Preparing data by sorting it on the group field before using Process When Group Changes makes a big difference in how efficiently the transformer can operate. This transformer is designed to react when the group value switches, so having all features belonging to the same group come in a single, consecutive block lets it process that entire group in one pass and then move on cleanly to the next. With data in this order, the transformer can minimize the amount of memory it needs to hold and reduce the frequency of state resets or boundary checks, which speeds up processing and lowers overhead.

If the input is not sorted by the grouping field, group changes can occur irregularly throughout the stream. The transformer then may need to buffer more features, manage more complex state, and frequently reset at unpredictable boundaries, all of which slow things down and increase resource use. That’s why pre-sorting is recommended for better performance when using Process When Group Changes.

Blocking mode and other options can influence performance in specific scenarios, but the general guidance for performance gains with this transformer is to pre-sort the data by the grouping field so groups are contiguous.