Utility node
for_each
For-each: split input into pieces, run a network on each (with detail attrib 'iteration', 'piece'), merge results. Per-piece cached.
Inputs
- 0geometry
- 1extra input
Parameters 4
networkstringNetwork run once per piece; subnet_input index 0 = piece, index 1 = extra input
piece_bymenuHow to split input 0 (prim attrib name/class, connectivity, every prim, every point, or N iterations)
iterationsintFor piece_by=count: run N times on the whole input (detail 'iteration' set)
merge_resultstoggleNo description.
Use it
Add it from the node menu (right-click empty graph space or Tab, then type for_each), ask the AI for it by name, or add it with a graph op from the API or MCP:
{"op":"add","type":"for_each","name":"my_for_each","params":{},"inputs":["src"]}Any parameter can also be an expression such as "=sin(F*0.1)". See Parameters & expressions.