Pack tuple

Combines multiple individual streams into a single tuple, preserving the order of elements.

Use this when

• A downstream component expects grouped data as a single structured unit and order matters more than names.
• You are zipping parallel outputs from sibling components for downstream processing.
• You want a lighter-weight alternative to Pack record when no field names are needed.

What it does

Consumes one message from each input stream per tick and emits them as a single tuple in input order. Inputs Segmentation, BoundingBox emit (Segmentation, BoundingBox).

Inputs

Type	Description
$ts...	One input stream per tuple position.

Outputs

Type	Description
($ts...)	A tuple with elements in order.

Parameters

This transformation takes no parameters.

Works best with

• Unpack tuple — natural inverse.
• Any downstream component that consumes a tuple as a single typed unit.

Caveats

• Position is significant — swapping input connections changes the tuple's semantics.
• Use Pack record when field names matter or reusability is important.