xad-rs 0.8.2

Exact automatic differentiation for Rust — forward-mode, reverse-mode, first- and second-order, with named variable support and a unified `Real` trait for mode-agnostic numerical code
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28

//! Forward-mode AD types.
//!
//! Forward mode augments every scalar with a *tangent* and propagates
//! both components together — the dual-number / k-jet construction. It
//! is the cheapest mode in this crate when the input dimension is small
//! and is what you want for a few directional derivatives or a full
//! Hessian on a moderate-sized input vector.
//!
//! - [`Jet1<T>`](freal::Jet1) — single-variable forward mode
//! - [`Jet1Vec`](dual::Jet1Vec) — multi-variable forward mode
//! - [`Jet2<T>`](dual2::Jet2) — second-order forward mode
//! - [`Jet2Vec`](dual2vec::Jet2Vec) — dense multi-variable second-order forward mode
//!
//! Each module also contains its named wrapper counterpart
//! (`NamedJet1`, `NamedJet1Vec`, `NamedJet2`).
//!
//! See also (theory): [`docs/theory/02-forward-mode-and-dual-numbers.md`](https://github.com/sercanatalik/xad-rs/blob/main/docs/theory/02-forward-mode-and-dual-numbers.md)
//! and [`docs/theory/04-second-order-and-k-jets.md`](https://github.com/sercanatalik/xad-rs/blob/main/docs/theory/04-second-order-and-k-jets.md).

pub mod jet1;
pub mod jet1vec;
pub mod jet2;
pub mod jet2vec;

pub use jet1::{Jet1, NamedJet1};
pub use jet1vec::{Jet1Vec, NamedJet1Vec};
pub use jet2::{Jet2, NamedJet2};
pub use jet2vec::Jet2Vec;