autodiff_rs 0.1.0

Automatic differentiation for Rust, with ndarray support
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
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69

use num::complex::Complex;
use num::rational::Ratio;
use num::{Integer, Num};
use std::num::Wrapping;

/// Compile-time calculation for what the gradient type should be based
/// on input and output types.
///
/// For most cases, the gradient type is the same as the output type.
/// However, for multi-parameter functions, like functions of arrays, this
/// is not the case
///
/// `<InputType as GradientType<OutputType>>::GradientType`
pub trait GradientType<OutputType> {
    /// The type of the gradient for a function with input type `Self` and output type `OutputType`
    type GradientType;
}

// macro to implement GradientType for a simple type
// where the gradient type is the same as the output type
macro_rules! impl_simple_types {
    ($($type:ty),*) => {
        $(
            // impl for values
            impl<T> GradientType<T> for $type {
                type GradientType = T;
            }
        )*
    };
}

// implement GradientType for all primitive types as well as complex numbers
impl_simple_types!(
    f32,
    f64,
    i8,
    i16,
    i32,
    i64,
    u8,
    u16,
    u32,
    u64,
    u128,
    i128,
    usize,
    isize,
    num::BigInt,
    num::BigUint,
    Complex<f32>,
    Complex<f64>
);

// generic impls
impl<T, U> GradientType<U> for Ratio<T>
where
    T: Integer + Clone,
    U: Num + Clone,
{
    type GradientType = Ratio<U>;
}

impl<T, U> GradientType<U> for Wrapping<T>
where
    T: Integer + Clone,
    U: Num + Clone,
{
    type GradientType = Wrapping<U>;
}