ida/
traits.rs

1//! Basic traits for problem specification
2
3use ndarray::prelude::*;
4
5/// Model specification
6pub trait ModelSpec: Clone {
7    type Scalar: num_traits::Float;
8    type Dim: Dimension;
9    fn model_size(&self) -> <Ix1 as Dimension>::Pattern;
10}
11
12pub trait Residual: ModelSpec {
13    /// Nonlinear residual function
14    //fn residual<'a, S>(&mut self, v: &'a mut ArrayBase<S, Ix1>) -> &'a mut ArrayBase<S, Ix1>
15    //where
16    //S: ndarray::DataMut<Elem = Self::Scalar>;
17
18    /// This function computes the problem residual for given values of the independent variable t, state vector y, and derivative ˙y. Arguments tt is the current value of the independent variable.
19    ///
20    /// # Arguments
21    ///
22    /// * `tt` is the current value of the independent variable.
23    /// * `yy` is the current value of the dependent variable vector, `y(t)`.
24    /// * `yp` is the current value of `y'(t)`.
25    /// * `rr` is the output residual vector `F(t, y, y')`.
26    ///
27    /// An IDAResFn function type should return a value of 0 if successful, a positive value if a recoverable error occurred (e.g. yy has an illegal value), or a negative value if a nonrecoverable error occurred. In the last case, the integrator halts. If a recoverable error occurred, the integrator will attempt to correct and retry.
28    fn res<S1, S2, S3>(
29        &self,
30        tt: Self::Scalar,
31        yy: ArrayBase<S1, Ix1>,
32        yp: ArrayBase<S2, Ix1>,
33        rr: ArrayBase<S3, Ix1>,
34    ) where
35        S1: ndarray::Data<Elem = Self::Scalar>,
36        S2: ndarray::Data<Elem = Self::Scalar>,
37        S3: ndarray::DataMut<Elem = Self::Scalar>;
38}
39
40pub trait Jacobian: ModelSpec {
41    /// This function computes the Jacobian matrix J of the DAE system (or an approximation to it)
42    ///
43    /// # Arguments
44    ///
45    /// * `tt` is the current value of the independent variable `t`.
46    /// * `cj` is the scalar in the system Jacobian, proportional to the inverse of the step size (α in Eq. (2.5)).
47    /// * `yy` is the current value of the dependent variable vector, `y(t)`.
48    /// * `yp` is the current value of `y'(t)`.
49    /// * `rr` is the current value of the residual vector `F(t, y, y')`.
50    /// * `jac` is the output (approximate) Jacobian matrix, `J = ∂F/∂y + cj ∂F/∂y'`.
51    ///
52    /// # Return value
53    ///
54    /// Should return 0 if successful, a positive value if a recoverable error occurred, or a
55    /// negative value if a nonrecoverable error occurred. In the case of a recoverable eror
56    /// return, the integrator will attempt to recover by reducing the stepsize, and hence changing α in
57    fn jac<S1, S2, S3, S4>(
58        &self,
59        tt: Self::Scalar,
60        cj: Self::Scalar,
61        yy: ArrayBase<S1, Ix1>,
62        yp: ArrayBase<S2, Ix1>,
63        rr: ArrayBase<S3, Ix1>,
64        jac: ArrayBase<S4, Ix2>,
65    ) where
66        S1: ndarray::Data<Elem = Self::Scalar>,
67        S2: ndarray::Data<Elem = Self::Scalar>,
68        S3: ndarray::Data<Elem = Self::Scalar>,
69        S4: ndarray::DataMut<Elem = Self::Scalar>;
70}
71
72pub trait Root: ModelSpec {
73    fn num_roots(&self) -> usize {
74        0
75    }
76
77    fn root<S1, S2, S3>(
78        &self,
79        t: Self::Scalar,
80        y: ArrayBase<S1, Ix1>,
81        yp: ArrayBase<S2, Ix1>,
82        gout: ArrayBase<S3, Ix1>,
83    ) where
84        S1: ndarray::Data<Elem = Self::Scalar>,
85        S2: ndarray::Data<Elem = Self::Scalar>,
86        S3: ndarray::DataMut<Elem = Self::Scalar>,
87    {
88    }
89}
90
91/// Core implementation for explicit schemes
92pub trait IdaProblem: Residual + Jacobian + Root {}
93
94impl<T> IdaProblem for T where T: Residual + Jacobian + Root {}
ida/traits.rs

ida/
traits.rs
