Struct worley_noise::WorleyNoise

pub struct WorleyNoise { /* fields omitted */ }

The base noise struct.

Uses permutation arrays for efficiency. Use the WorleyNoise::permutate function to randomize it / adjust the size.

Caches already sampled values.

Sampling individual points is relatively slow due to the internal locks. Use the WorleyNoise::values function if sampling multiple points instead as it's concurrent.

Methods

impl WorleyNoise

pub const DEFAULT_RADIUS: u16

DEFAULT_RADIUS: u16 = 1

pub const DEFAULT_PERMUTATION_BITS: usize

DEFAULT_PERMUTATION_BITS: usize = 10

pub const DEFAULT_DENSITY: f64

DEFAULT_DENSITY: f64 = 3.0

pub const DEFAULT_CACHE_CAPACITY: usize

DEFAULT_CACHE_CAPACITY: usize = 2500

pub fn new() -> Self

Creates a new noise struct with random permutation arrays.

Uses the default cache capacity of 2500.

pub fn with_cache_capacity(capacity: usize) -> Self

Initializes the struct with the specified cache capacity.

pub fn permutate(&mut self, permutation_table_bit_length: usize)

Calls WorleyNoise::permutate_seeded with a random seed.

pub fn permutate_seeded(     &mut self,     permutation_table_bit_length: usize,     seed: u128 )

Randomizes the internal permutation arrays.

Generates permutation arrays with two to the power of permutation_table_bit_length entries.

This operation is fairly slow.

pub fn set_density(&mut self, density: f64)

Sets the feature point density.

Might be slow since it precomputes a lot of stuff.

Default is 3.0.

pub fn set_distance_function<F>(&mut self, function: F) where     F: Fn(f64, f64) -> f64 + Sync + 'static,

Sets the function to calculate the distance between feature points.

Default is the squared Euclidean distance.

pub fn set_value_function<F>(&mut self, function: F) where     F: Fn(Vec<f64>) -> f64 + Sync + 'static,

Sets the function to pick the final value from the nearby feature points.

The values are in no particular order.

Default is the minimum value.

pub fn set_radius(&mut self, radius: u16)

Specifies how many adjacent cells should be included in the value calculation.

A higher radius reduces the amount of possible errors but slows down sampling of individual points.

Default is 1. That means it includes the feature points of the surrounding cells.

pub fn value(&self, x: f64, y: f64) -> f64

Calculates the noise value for the given point.

pub fn values(&self, points: &Vec<(f64, f64)>) -> Vec<f64>

Calculates the noise values for the given points in parallel.