I'm starting to learn rust, so let me know if there is a more proper way I should be approaching this problem, but I'm looking for similar performance of numpy from python when doing element wise operations on vectors. It looks like the ndarray crate is a good fit, but I'm having a hard time turning a Vec into a ndarray. Starting with:
let mut A:Vec<f32> = vec![];
How would I do something like:
let B = array!(A);
after filling in the vector?
I didn't see a way to cast the vector to an ndarray like I would in python with lists. Do I really have to just use a for loop and set each index individually to an empty ndarray?
You can use from_vec to create an ndarray from a vector or from_shape_vec to create a vector of a custom dimension.
// You can use from_vec
let one_dim_vector = Array::from_vec(vec![1, 2, 3, 4]);
println!("Vector created from_vec:\n{:#?}", one_dim_vector);
Or you can create it from an iterable:
let one_dim_vector_from_iterable = Array::from_iter(1..5);
println!(
"Vector created from_iter:\n{:#?}",
one_dim_vector_from_iterable
);
You can also create it for a custom dimension passing a tuple of (rows, columns):
let vector_of_custom_dimension = Array::from_shape_vec((3, 2), vec![1., 2., 3., 4., 5., 6.]);
match vector_of_custom_dimension {
Ok(vector) => println!("Vector created from_shape_vec:\n{:#?}", vector),
Err(error) => println!("Error: {:#?}", error),
}
This will output:
Vector created from_vec: [1, 2, 3, 4]
Vector created from_iter: [1, 2, 3, 4]
Vector created from_shape_vec:
[[1.0, 2.0],
[3.0, 4.0]],