How to Map and Fit Point Numbers to Custom Ranges in Houdini VEX

Remapping Point Indices in Houdini VEX

When working with geometry in Houdini, normalizing point numbers (@ptnum) against the total point count (@numpt) is a foundational technique for creating procedural gradients, controlling falloffs, and driving downstream attributes.

The Common Pitfall: Integer Division & Range Mappings

In VEX, @ptnum and @numpt are integers. Dividing two integers can cause precision errors if not cast to floating-point numbers. Additionally, manually scaling fractions can lead to syntax issues and calculation errors.

Here is how to properly normalize and remap point ranges in a Point Wrangle.

1. Basic Normalization (0 to 1)

To map point indices linearly from 0.0 (first point) to 1.0 (last point):

// Cast to float to avoid integer division truncation
float normalized = (float)@ptnum / (float)(@numpt - 1);

// Assign to a custom point attribute
@my_val = normalized;

2. Inverting the Range (1 to 0)

To make the first point @ptnum = 0 output 1.0 and the last point output 0.0:

float normalized = (float)@ptnum / (float)(@numpt - 1);
@my_val = 1.0 - normalized;

3. The Recommended Method: Using fit() or fit01()

Instead of manually multiplying and adding offsets, VEX provides built-in fitting functions that handle the interpolation safely and cleanly.

Mapping to an Arbitrary Range (e.g., 0.1 to 1.0)

// Using fit01 to remap a normalized 0-1 value to 0.1 - 1.0
float normalized = (float)@ptnum / (float)(@numpt - 1);
@my_val = fit01(normalized, 0.1, 1.0);

Alternative Direct fit() Syntax

// Remap directly from index domain [0, @numpt - 1] to target domain [0.1, 1.0]
@my_val = fit((float)@ptnum, 0.0, (float)(@numpt - 1), 0.1, 1.0);

Verification Example (5 Total Points)

When @numpt = 5, indices run from 0 to 4: