This is a very fast Perlin noise library, mostly in Go with Python bindings, that I developed upon trying out a few other available Python libraries and realized that none were fast enough for my project requirements. Algorithms are from the libnoise dotnet project which I ported to Go.
To install fast_perlin_noise,
pip3 install fast-perlin-noiseWheels are automatically built for Windows, Linux, and macOS for x86-64 and ARM64 architectures. If a wheel does not exist for your platform, you will require a Go compiler to complete the installation.
fast_perlin_noise has very limited dependencies: only numpy is required!
fast_perlin_noise has a very beginner friendly Python interface, with optional advanced use.
from fast_perlin_noise import PerlinNoise
import numpy as np
import matplotlib.pyplot as plt
noise_generator: PerlinNoise = PerlinNoise(width=256, height=256)
noise_image: np.ndarray = noise_generator.generate_noise_matrix()
plt.imshow(noise_image) # View the resulting noise
You can run and look at example/example.py to see this in action.
fast_perlin_noise currently outputs a matrix (an ndarray with the shape m, n) of noise of which the values range from 0.0 to 1.0.
For more advanced use, many parameters can be tuned to adjust the resulting noise.
Perlin noise can be generated using the generate_noise_matrix function.
| Parameter | Type | Description |
|---|---|---|
| width | uint | Width of resultant matrix. |
| height | uint | Height of resultant matrix. |
| persistence | float | Intensity falloff coefficient of subsequent noise layers. |
| numLayers | uint | Number of simplex noise layers to use. |
| roughness | float | Frequency increase coefficient for subsequent noise layers. |
| baseRoughness | float | Initial frequency for noise |
| strength | float | Scalar multiplier for noise values |
| randomSeed | float | Define the random seed to be used |
Behaviour of the noise can be changed by changing parameters to the PerlinNoise interface.
Settings include width and height to determine the dimensions of the resulting matrix. baseRoughness and roughness to control the frequency and frequency falloff.
persistence to change the impact of subsequent layers (low value results in a "softer" look). numLayers controls how many layers of noise will be used (more layers results in more complex, structured noise).
strength is a simple scalar multiplier to the matrix (control intensity of noise). Lastly, randomSeed can be changed to change the seed of the noise (fast_perlin_noise is deterministic when randomSeed is known).