autodj 0.5.3

Automatic Differentiation Library
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

//! [`crate::solid::DualNumber`] for single variable differentiations

use crate::fluid::{Dual, Value};

/// Uni-variate dual number
///
///```
/// use autodj::prelude::single::*;
/// let x0 : DualF64 = 1.0.into();
/// let x  : DualF64 = 3.0.into_variable();
/// let f = (x - x0).powf(2.0);
/// assert_eq!(f.value(), &4.);
/// assert_eq!(f.dual(),  &4.);
/// ```
pub type DualNumber<V> = super::DualNumber<V, V>;

/// Single [`f64`] variable
pub type DualF64 = DualNumber<f64>;

/// Single [`f32`] variable
pub type DualF32 = DualNumber<f32>;

// TODO: is it generalizable for multivariate ?
/// Create an independent variable from a value
pub trait IntoVariable: Value {
    /// Create an independent variable from a value
    fn into_variable(self) -> DualNumber<Self> {
        Dual::new(self, Self::one())
    }
}

impl<V: Value> IntoVariable for V {}