Crates.io | dace |
lib.rs | dace |
version | 0.2.1 |
source | src |
created_at | 2023-01-08 13:52:42.885269 |
updated_at | 2024-04-07 21:52:10.670316 |
description | Rust wrapper of DACE, the Differential Algebra Computational Toolbox. |
homepage | |
repository | https://github.com/giovannipurpura/dace-rs/ |
max_upload_size | |
id | 753700 |
size | 598,151 |
Rust wrapper of DACE, the Differential Algebra Computational Toolbox.
DACE-RS is a Rust wrapper of DACE, the Differential Algebra Computational Toolbox (https://github.com/dacelib/dace).
You can find further details on Differential Algebra and its applications on that web page.
DACE-RS can be used in your project by including it as a Cargo dependency.
Add this to your Cargo.toml
:
[dependencies]
dace = "0.2"
If you need to use the AlgebraicVector<DA>::invert()
function,
you need to include ndarray-linalg
and specify a LAPACK binding
(see: https://github.com/rust-ndarray/ndarray-linalg).
Example:
[dependencies]
dace = "0.2"
ndarray-linalg = { version = "0.16", features = ["openblas-static"] }
This is needed also to run tests, e.g.:
cargo test --features=ndarray-linalg/openblas-static
CMake and a C compiler must be installed in the system to build the DACE Core library.
The original DACE C++ tutorials have been translated to Rust and are available in the examples folder: https://github.com/giovannipurpura/dace-rs/tree/master/examples
This is a quick example for basic usage:
use dace::*; // import all DACE elements
fn main() {
// initialize DACE with order 10 and 3 variables
DA::init(10, 3);
// assign the three variables to x, y, z -- notice that integers are used here!
let (x, y, z): (DA, DA, DA) = (da!(1), da!(2), da!(3));
// create also some constants as DA objects -- notice that floats are used here!
let (a, b, c): (DA, DA, DA) = (da!(1.0), da!(2.0), da!(3.0));
// compute a * sin(x) + b * cos(y) + c * tan(z)
let v1: DA = &a * x.sin() + &b * y.cos() + &c * z.tan();
// print the resulting DA variable
println!("{v1}");
// do the same without using the DA constants a, b, c
let v2: DA = 1.0 * x.sin() + 2.0 * y.cos() + 3.0 * z.tan();
// check that we got the same result
println!("v1 == v2: {}", v1 == v2);
// try also with AlgebraicVector<DA> and AlgebraicVector<f64>
let xyz: AlgebraicVector<DA> = darray![x.sin(), y.cos(), z.tan()];
let abc: AlgebraicVector<f64> = darray![1.0, 2.0, 3.0];
let v3: DA = xyz.dot(&abc);
// check that we got the same result
println!("v1 == v3: {}", v1 == v3);
// try also with AlgebraicMatrix<DA> and AlgebraicMatrix<f64>
let xyz: AlgebraicMatrix<DA> = darray![[x.sin(), y.cos(), z.tan()]];
let abc: AlgebraicMatrix<f64> = darray![[1.0], [2.0], [3.0]];
let v4: AlgebraicMatrix<DA> = xyz.dot(&abc);
// check that we got the same result
println!("v1 == v4: {}", v1 == v4[(0, 0)]);
}