Skip to content

Releases: Kromey/fast_poisson

v1.0.2

26 Mar 17:20
v1.0.2
5c8273c
Compare
Choose a tag to compare
chore: Release fast_poisson version 1.0.2

v1.0.0

28 Apr 23:32
v1.0.0
a5dce89
Compare
Choose a tag to compare

PRNG algorithm can now be selected directly by the user, or left at the default.
Bugfixes and overhauling some internals.

v0.5.2

12 Jul 00:23
v0.5.2
4cbf742
Compare
Choose a tag to compare

Revert Rust edition to 2018 (changed to 2021 in v0.5.1), as 2021 edition is unstable in our MSRV of 1.51.0.

v0.5.1

11 Jul 23:13
v0.5.1
637aa5d
Compare
Choose a tag to compare

Builder-style functions return a mutable reference, allowing them to be chained as intended.

v0.4.0

26 Apr 20:04
Compare
Choose a tag to compare

This version brings several new features and bug fixes:

  • Added new single_precision feature flag to support use cases that require f32 instead of f64
  • Added new small_rng feature flag to allow choosing a PRNG with a smaller state overhead
  • Implement PoissonIter as a FusedIterator to allow more efficient fuse() operations
  • Various bug fixes, including fixing the lookup grid size and the point sampling algorithm

This version also changes the licensing to either of Apache 2.0 or MIT licenses.

Documentation fixes

06 Apr 19:48
Compare
Choose a tag to compare
v0.3.1

Bump version

Conversion to Vec

06 Apr 19:11
Compare
Choose a tag to compare

Poisson objects can now be converted directly into Vec; see the docs for details.

Internally, non-seeded Poisson distributions have been corrected to use the recommended method for constructing non-deterministic PRNGs.

N-Dimensional Distributions

29 Mar 20:16
Compare
Choose a tag to compare
  • Added support for n-dimensional Poisson distributions
  • Dropped direct instantiation of the Poisson struct for the builder pattern
  • Leveraged Rust's const generics to return n-dimensional points as arrays

First release

25 Mar 23:50
Compare
Choose a tag to compare

Provides an iterator that lazily evaluates the Poisson disk distribution using Bridson's algorithm.