Prune

Prune removes existing branches when certain conditions are met.
It evaluates attributes on internodes and marks them for removal when thresholds or rules are satisfied.

Unlike one-shot conditions, Prune can re-evaluate over time, which makes it effective when attribute values change after creation — for example, branches that should only die once they reach a certain height or age.

Concept

• Conditional removal — check an attribute against a constant, another attribute, or a mapping.
• Cycles — each internode is tested repeatedly during its lifetime, so pruning is not limited to a single chance.
• Delay — growth can proceed for some time before pruning begins, allowing controlled staging.
• Iterative feedback — combines well with Wrangles or Mappings that update attributes across iterations.

⚠️ Note: Natsura currently has limited support for attributes that evolve over time on already-created geometry, except when using Wrangle. Wrangles can inject dynamic logic that makes Prune more powerful.

Inputs

Input 0 — Graph
The incoming Apex graph.

Output — Graph
The same graph with pruned internodes removed, ready for downstream evaluation or Simulate.

Parameters

• Compare Attribute (graph_attribute_a)
  Attribute to test (e.g. height, age, width).
• Condition (graph_threshold_operation)
  Operator (>=, <=, ==, etc.).
• Test Value (gate_value_b)
  Constant value to compare against.
• Attribute / Map Mode (graph_type_b)
  Instead of a constant, you can compare against another attribute or a mapping chain.
• Snippet Condition (graph_snippet_condition)
  Optional VEX snippet for custom gating logic.
• Test Lifetime (cycles)
  Number of times the prune condition is re-evaluated.
• Start Delay (delay)
  Frames/steps to wait before pruning begins.

Attributes

Prune typically uses attributes already present in the graph, such as:

• height, age, generation
• u, repeat_index
• Any custom attributes created by Wrangle or Mapping chains

Workflow

1. Insert a Prune node after growth logic (e.g. after Grow or inside a Repeat).
2. Choose an attribute and condition.
3. Set Test Lifetime to give branches multiple evaluation chances.
4. Use Start Delay to allow some growth before pruning begins.

How-tos

Kill tall branches

/// Remove when taller than 5 units
Compare Attribute: height
Condition: >=
Test Value: 5

Density-based cleanup with Wrangle

// In Wrangle: add a density score
f@local_density = nearpoints(0, @P, 0.5);

Then in Prune, set:

• Compare Attribute: local_density
• Condition: >
• Test Value: 4

Branches in overcrowded areas are removed over time.

See also

• Grow – generates internodes to be pruned
• Repeat – time-based iteration, useful before pruning
• Wrangle – inject custom attributes for pruning
• Switch – alternative gating based on conditions