Under what condition does the Process When Group Changes (Advanced) setting in group-based transformers help performance?

Under what condition does the Process When Group Changes (Advanced) setting in group-based transformers help performance?

• A. Features are geographically sorted using a nearest neighbor method.
• B. Features are less than 1 mb each.
• C. Features pre-sorted into their groups.
• D. Features are sorted by a universally unique identifier (UUID)

Answer: (C)

The setting is most effective when features arrive already organized by their group, so the transformer can treat each group as a unit and only reset its processing when the group changes. When features are pre-sorted into their groups, the stream naturally contains long runs of features from the same group, allowing the transformer to batch work per group and avoid repeated reinitialization for every feature. This reduces overhead and speeds up processing. The other conditions—sorting by geography, small feature size, or sorting by a UUID—don’t align with group boundaries or batching behavior, so they don’t provide the same performance benefit.

The setting is most effective when features arrive already organized by their group, so the transformer can treat each group as a unit and only reset its processing when the group changes. When features are pre-sorted into their groups, the stream naturally contains long runs of features from the same group, allowing the transformer to batch work per group and avoid repeated reinitialization for every feature. This reduces overhead and speeds up processing. The other conditions—sorting by geography, small feature size, or sorting by a UUID—don’t align with group boundaries or batching behavior, so they don’t provide the same performance benefit.