This package is the counterpart of AbstractArray interface, but for GPU array types. It provides functionality and tooling to speed-up development of new GPU array types. This package is not intended for end users; instead, you should use one of the packages that builds on GPUArrays.jl, such as CUDA.jl, oneAPI.jl or AMDGPU.jl.

This package provides tools to express a design pattern for dealing with large/ nested structures, as in machine learning and optimisation. For large machine learning models it can be cumbersome or inefficient to work with parameters as one big, flat vector, and structs help in managing complexity; but it is also desirable to easily operate over all parameters at once, e.g. for changing precision or applying an optimiser update step.

This library provides tools for working with Julia code and expressions. This includes a template-matching system and code-walking tools that let you do deep transformations of code.

This package supports SI units and also many other unit system.

A Julia package for evaluating distances(metrics) between vectors. This package also provides optimized functions to compute column-wise and pairwise distances, which are often substantially faster than a straightforward loop implementation.

IteratorInterfaceExtensions defines a small number of extensions to the iterator interface.

This package implements various 3D rotation parameterizations and defines conversions between them. At their heart, each rotation parameterization is a 3×3 unitary (orthogonal) matrix (based on the StaticArrays.jl package), and acts to rotate a 3-vector about the origin through matrix-vector multiplication.

This package provides a summary of available CPU features in Julia.

OffsetArrays.jl provides Julia users with arrays that have arbitrary indices, similar to those found in some other programming languages like Fortran.

Quaternions are best known for their suitability as representations of 3D rotational orientation. They can also be viewed as an extension of complex numbers.

This package provides these irrational constants:

• twoπ = 2π

• fourπ = 4π

• halfπ = π / 2

• quartπ = π / 4

• invπ = 1 / π

• twoinvπ = 2 / π

• fourinvπ = 4 / π

• inv2π = 1 / (2π)

• inv4π = 1 / (4π)

• sqrt2 = √2

• sqrt3 = √3

• sqrtπ = √π

• sqrt2π = √2π

• sqrt4π = √4π

• sqrthalfπ = √(π / 2)

• invsqrt2 = 1 / √2

• invsqrtπ = 1 / √π

• invsqrt2π = 1 / √2π

• loghalf = log(1 / 2)

• logtwo = log(2)

• logten = log(10)

• logπ = log(π)

• log2π = log(2π)

• log4π = log(4π)

StackViews provides only one array type: StackView. There are multiple ways to understand StackView:

• inverse of eachslice

• cat variant

• view object

• lazy version of repeat special case

This package provides a wrapper around ImageMagick version 6. It was split off from Images.jl to make image I/O more modular.

ReverseDiff.jl is a fast and compile-able tape-based reverse mode AD, that implements methods to take gradients, Jacobians, Hessians, and higher-order derivatives of native Julia functions (or any callable object, really).

This package provides a function extrapolate that extrapolates a given function f(x) to f(x0), evaluating f only at a geometric sequence of points > x0 (or optionally < x0). The key algorithm is Richardson extrapolation using a Neville–Aitken tableau, which adaptively increases the degree of an extrapolation polynomial until convergence is achieved to a desired tolerance (or convergence stalls due to e.g. floating-point errors). This allows one to obtain f(x0) to high-order accuracy, assuming that f(x0+h) has a Taylor series or some other power series in h.

This package provides definitions for most of the primary types and functions in StaticArrays.jl. This enables downstream packages to implement new methods on these types without depending on the entirety of StaticArrays.jl.

Various special functions based on log and exp moved from StatsFuns.jl into a separate package, to minimize dependencies. These functions only use native Julia code, so there is no need to depend on librmath or similar libraries.

DeepDiffs.jl provides the deepdiff function, which finds and displays differences (diffs) between Julia data structures. It supports Vectors, Dicts, and Strings. When diffing dictionaries where values associated with a particular key may change, deepdiff will recurse into value to provide a more detailed diff.

This package provides a namespace for data-related generic function definitions to solve the optional dependency problem; packages wishing to share and/or extend functions can avoid depending directly on each other by moving the function definition to DataAPI.jl and each package taking a dependency on it.

This package provides definitions for common functions that are useful for symbolic expression manipulation in Julia. Its purpose is to provide a shared interface between various symbolic programming packages, for example SymbolicUtils.jl, Symbolics.jl, and Metatheory.jl.

The purpose of this package is partly to extend linear algebra functionality in base to cover generic element types, e.g. BigFloat and Quaternion, and partly to be a place to experiment with fast linear algebra routines written in Julia (except for optimized BLAS).

This package provides a collection of tools for metaprogramming on Julia Expr, the meta programming standard library for MLStyle.

This package takes a string or buffer containing Julia code, performs lexical analysis and returns a stream of tokens.

This package is designed to help in testing ChainRulesCore.frule and ChainRulesCore.rrule methods. The main entry points are ChainRulesTestUtils.frule_test, ChainRulesTestUtils.rrule_test, and ChainRulesTestUtils.test_scalar. Currently this is done via testing the rules against numerical differentiation (using FiniteDifferences.jl).

ChainRulesTestUtils.jl is separated from ChainRulesCore.jl so that it can be a test-only dependency, allowing it to have potentially heavy dependencies, while keeping ChainRulesCore.jl as light-weight as possible.