Perlin Noise

1.14 verified Sat May 09 auth: no python

A Python implementation of Perlin Noise with support for unlimited coordinate space. Current version 1.14, requires Python >=3.9,<3.13. Released under MIT license. Active development with periodic updates.

pip install perlin-noise

Common errors

error ModuleNotFoundError: No module named 'perlin_noise' ↓

cause Package name uses hyphen (perlin-noise) but import requires underscore.

fix

Install with 'pip install perlin-noise' and import as 'from perlin_noise import PerlinNoise'.

error TypeError: argument of type 'int' is not iterable ↓

cause Passing a single integer as coordinate instead of a list/tuple.

fix

Use noise([x]) for 1D or noise([x, y]) for 2D.

error AttributeError: module 'perlin_noise' has no attribute 'PerlinNoise2D' ↓

cause Using deprecated class name removed in version 1.12+.

fix

Update code to use 'from perlin_noise import PerlinNoise' and pass coordinates as list.

Warnings

breaking In version 1.0, the seed parameter was named 'seed' but was renamed to 'random_seed' in a later release (<1.10). If using older code, update to 'seed' or check the version. ↓

fix Use 'seed' for current versions; or pin to older versions if needed.

gotcha The noise output values are not normalized to [0,1] by default; they typically range around [-1,1]. Some users expect a different range. ↓

fix Scale output manually if needed: e.g., (value + 1) / 2 for [0,1].

deprecated The 'PerlinNoise2D' and 'PerlinNoise3D' classes (from older versions) are deprecated; use the general 'PerlinNoise' class with appropriate coordinate dimensions. ↓

fix Use 'from perlin_noise import PerlinNoise' and pass [x,y] or [x,y,z].

Imports

PerlinNoise
```
from perlin_noise import PerlinNoise
```
Correct import path (note the underscore in the module name).
PerlinNoise
wrong
```
from perlin-noise import PerlinNoise
```
correct
```
from perlin_noise import PerlinNoise
```
Hyphens in package name are not valid in Python import statements; use underscores.

Quickstart

Create a seeded PerlinNoise object and generate a 2D noise map using a grid of coordinates.

from perlin_noise import PerlinNoise
import numpy as np

# Create a noise object with seed 10
noise = PerlinNoise(octaves=4, seed=10)

# Generate 2D noise (e.g., 10x10 grid)
freq = 16  # Adjust for desired detail
xp, yp = np.mgrid[0:10, 0:10] / freq
pic = np.array([[noise([x, y]) for y in yp[0]] for x in xp[:,0]])

print(pic.shape)  # (10,10)