srenblad/nalgebra
1
0
Fork 0
forked from M-Labs/nalgebra
Code Pull Requests Activity

Fix the return type of convolve_same to match the documentation.

Browse Source
This commit is contained in:
sebcrozet 2019-03-31 17:03:02 +02:00
parent ae4afa3d2c
commit bb06701eff
2 changed files with 37 additions and 37 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

28
src/linalg/convolution.rs
View File

@ -7,14 +7,14 @@ use crate::storage::Storage;
use crate::{zero, RealField, Vector, VectorN, U1}; use crate::{zero, RealField, Vector, VectorN, U1};
impl<N: RealField, D1: Dim, S1: Storage<N, D1>> Vector<N, D1, S1> { impl<N: RealField, D1: Dim, S1: Storage<N, D1>> Vector<N, D1, S1> {
/// Returns the convolution of the target vector and a kernel /// Returns the convolution of the target vector and a kernel.
/// ///
/// # Arguments /// # Arguments
/// ///
/// * `kernel` - A Vector with size > 0 /// * `kernel` - A Vector with size > 0
/// ///
/// # Errors /// # Errors
/// Inputs must statisfy `vector.len() >= kernel.len() > 0`. /// Inputs must satisfy `vector.len() >= kernel.len() > 0`.
/// ///
pub fn convolve_full<D2, S2>( pub fn convolve_full<D2, S2>(
&self, &self,
@ -53,7 +53,8 @@ impl<N: RealField, D1: Dim, S1: Storage<N, D1>> Vector<N, D1, S1> {
} }
conv conv
} }
/// Returns the convolution of the target vector and a kernel /// Returns the convolution of the target vector and a kernel.
///
/// The output convolution consists only of those elements that do not rely on the zero-padding. /// The output convolution consists only of those elements that do not rely on the zero-padding.
/// # Arguments /// # Arguments
/// ///
@ -61,10 +62,9 @@ impl<N: RealField, D1: Dim, S1: Storage<N, D1>> Vector<N, D1, S1> {
/// ///
/// ///
/// # Errors /// # Errors
/// Inputs must statisfy `self.len() >= kernel.len() > 0`. /// Inputs must satisfy `self.len() >= kernel.len() > 0`.
/// ///
pub fn convolve_valid<D2, S2>(&self, kernel: Vector<N, D2, S2>, pub fn convolve_valid<D2, S2>(&self, kernel: Vector<N, D2, S2>) -> VectorN<N, DimDiff<DimSum<D1, U1>, D2>>
) -> VectorN<N, DimDiff<DimSum<D1, U1>, D2>>
where where
D1: DimAdd<U1>, D1: DimAdd<U1>,
D2: Dim, D2: Dim,
@ -90,20 +90,20 @@ impl<N: RealField, D1: Dim, S1: Storage<N, D1>> Vector<N, D1, S1> {
conv conv
} }
/// Returns the convolution of the targetvector and a kernel /// Returns the convolution of the target vector and a kernel.
///
/// The output convolution is the same size as vector, centered with respect to the ‘full’ output. /// The output convolution is the same size as vector, centered with respect to the ‘full’ output.
/// # Arguments /// # Arguments
/// ///
/// * `kernel` - A Vector with size > 0 /// * `kernel` - A Vector with size > 0
/// ///
/// # Errors /// # Errors
/// Inputs must statisfy `self.len() >= kernel.len() > 0`. /// Inputs must satisfy `self.len() >= kernel.len() > 0`.
pub fn convolve_same<D2, S2>(&self, kernel: Vector<N, D2, S2>) -> VectorN<N, DimMaximum<D1, D2>> pub fn convolve_same<D2, S2>(&self, kernel: Vector<N, D2, S2>) -> VectorN<N, D1>
where where
D1: DimMax<D2>, D2: Dim,
D2: DimMax<D1, Output = DimMaximum<D1, D2>>,
S2: Storage<N, D2>, S2: Storage<N, D2>,
DefaultAllocator: Allocator<N, DimMaximum<D1, D2>>, DefaultAllocator: Allocator<N, D1>,
{ {
let vec = self.len(); let vec = self.len();
let ker = kernel.len(); let ker = kernel.len();
@ -112,8 +112,7 @@ impl<N: RealField, D1: Dim, S1: Storage<N, D1>> Vector<N, D1, S1> {
panic!("convolve_same expects `self.len() >= kernel.len() > 0`, received {} and {} respectively.",vec,ker); panic!("convolve_same expects `self.len() >= kernel.len() > 0`, received {} and {} respectively.",vec,ker);
} }
let result_len = self.data.shape().0.max(kernel.data.shape().0); let mut conv = VectorN::zeros_generic(self.data.shape().0, U1);
let mut conv = VectorN::zeros_generic(result_len, U1);
for i in 0..vec { for i in 0..vec {
for j in 0..ker { for j in 0..ker {
@ -125,6 +124,7 @@ impl<N: RealField, D1: Dim, S1: Storage<N, D1>> Vector<N, D1, S1> {
conv[i] += val * kernel[ker - j - 1]; conv[i] += val * kernel[ker - j - 1];
} }
} }
conv conv
} }
} }

2
tests/linalg/convolution.rs
View File

@ -11,7 +11,7 @@ use std::panic;
#[test] #[test]
fn convolve_same_check(){ fn convolve_same_check(){
// Static Tests // Static Tests
let actual_s = Vector4::from_vec(vec![1.0,4.0,7.0,10.0]); let actual_s = Vector4::new(1.0, 4.0, 7.0, 10.0);
let expected_s = Vector4::new(1.0, 2.0, 3.0, 4.0).convolve_same(Vector2::new(1.0, 2.0)); let expected_s = Vector4::new(1.0, 2.0, 3.0, 4.0).convolve_same(Vector2::new(1.0, 2.0));
assert!(relative_eq!(actual_s, expected_s, epsilon = 1.0e-7)); assert!(relative_eq!(actual_s, expected_s, epsilon = 1.0e-7));