work/smartcore
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Merge branch 'development' into prdct-prb

Browse Source
This commit is contained in:
Lorenzo
2023-01-27 10:42:48 +00:00
committed by GitHub
parent dae556776c c7353d0b57
commit 97604a2d83
47 changed files with 146 additions and 222 deletions
Download Patch File Download Diff File Expand all files Collapse all files
src/algorithm/neighbour/fastpair.rs
+9 -13
View File
@@ -260,8 +260,8 @@ mod tests_fastpair {
let distances = fastpair.distances;
let neighbours = fastpair.neighbours;
assert!(distances.len() != 0);
assert!(neighbours.len() != 0);
assert!(!distances.is_empty());
assert!(!neighbours.is_empty());
assert_eq!(10, neighbours.len());
assert_eq!(10, distances.len());
@@ -276,17 +276,13 @@ mod tests_fastpair {
// We expect an error when we run `FastPair` on this dataset,
// becuase `FastPair` currently only works on a minimum of 3
// points.
let _fastpair = FastPair::new(&dataset);
let fastpair = FastPair::new(&dataset);
assert!(fastpair.is_err());
match _fastpair {
Err(e) => {
let expected_error =
Failed::because(FailedError::FindFailed, "min number of rows should be 3");
assert_eq!(e, expected_error)
}
_ => {
assert!(false);
}
if let Err(e) = fastpair {
let expected_error =
Failed::because(FailedError::FindFailed, "min number of rows should be 3");
assert_eq!(e, expected_error)
}
}
@@ -582,7 +578,7 @@ mod tests_fastpair {
};
for p in dissimilarities.iter() {
if p.distance.unwrap() < min_dissimilarity.distance.unwrap() {
min_dissimilarity = p.clone()
min_dissimilarity = *p
}
}
src/cluster/dbscan.rs
+1 -1
View File
@@ -511,6 +511,6 @@ mod tests {
.and_then(|dbscan| dbscan.predict(&x))
.unwrap();
println!("{:?}", labels);
println!("{labels:?}");
}
}
src/cluster/kmeans.rs
+2 -2
View File
@@ -498,8 +498,8 @@ mod tests {
let y: Vec<usize> = kmeans.predict(&x).unwrap();
for i in 0..y.len() {
assert_eq!(y[i] as usize, kmeans._y[i]);
for (i, _y_i) in y.iter().enumerate() {
assert_eq!({ y[i] }, kmeans._y[i]);
}
}
src/dataset/boston.rs
+1 -1
View File
@@ -31,7 +31,7 @@ use crate::dataset::Dataset;
pub fn load_dataset() -> Dataset<f32, f32> {
let (x, y, num_samples, num_features) = match deserialize_data(std::include_bytes!("boston.xy"))
{
Err(why) => panic!("Can't deserialize boston.xy. {}", why),
Err(why) => panic!("Can't deserialize boston.xy. {why}"),
Ok((x, y, num_samples, num_features)) => (x, y, num_samples, num_features),
};
src/dataset/breast_cancer.rs
+1 -1
View File
@@ -33,7 +33,7 @@ use crate::dataset::Dataset;
pub fn load_dataset() -> Dataset<f32, u32> {
let (x, y, num_samples, num_features) =
match deserialize_data(std::include_bytes!("breast_cancer.xy")) {
Err(why) => panic!("Can't deserialize breast_cancer.xy. {}", why),
Err(why) => panic!("Can't deserialize breast_cancer.xy. {why}"),
Ok((x, y, num_samples, num_features)) => (
x,
y.into_iter().map(|x| x as u32).collect(),
src/dataset/diabetes.rs
+1 -1
View File
@@ -26,7 +26,7 @@ use crate::dataset::Dataset;
pub fn load_dataset() -> Dataset<f32, u32> {
let (x, y, num_samples, num_features) =
match deserialize_data(std::include_bytes!("diabetes.xy")) {
Err(why) => panic!("Can't deserialize diabetes.xy. {}", why),
Err(why) => panic!("Can't deserialize diabetes.xy. {why}"),
Ok((x, y, num_samples, num_features)) => (
x,
y.into_iter().map(|x| x as u32).collect(),
src/dataset/digits.rs
+1 -1
View File
@@ -16,7 +16,7 @@ use crate::dataset::Dataset;
pub fn load_dataset() -> Dataset<f32, f32> {
let (x, y, num_samples, num_features) = match deserialize_data(std::include_bytes!("digits.xy"))
{
Err(why) => panic!("Can't deserialize digits.xy. {}", why),
Err(why) => panic!("Can't deserialize digits.xy. {why}"),
Ok((x, y, num_samples, num_features)) => (x, y, num_samples, num_features),
};
src/dataset/iris.rs
+1 -1
View File
@@ -22,7 +22,7 @@ use crate::dataset::Dataset;
pub fn load_dataset() -> Dataset<f32, u32> {
let (x, y, num_samples, num_features): (Vec<f32>, Vec<u32>, usize, usize) =
match deserialize_data(std::include_bytes!("iris.xy")) {
Err(why) => panic!("Can't deserialize iris.xy. {}", why),
Err(why) => panic!("Can't deserialize iris.xy. {why}"),
Ok((x, y, num_samples, num_features)) => (
x,
y.into_iter().map(|x| x as u32).collect(),
src/dataset/mod.rs
+1 -1
View File
@@ -78,7 +78,7 @@ pub(crate) fn serialize_data<X: Number + RealNumber, Y: RealNumber>(
.collect();
file.write_all(&y)?;
}
Err(why) => panic!("couldn't create {}: {}", filename, why),
Err(why) => panic!("couldn't create {filename}: {why}"),
}
Ok(())
}
src/decomposition/pca.rs
+9 -11
View File
@@ -231,8 +231,7 @@ impl<T: Number + RealNumber, X: Array2<T> + SVDDecomposable<T> + EVDDecomposable
if parameters.n_components > n {
return Err(Failed::fit(&format!(
"Number of components, n_components should be <= number of attributes ({})",
n
"Number of components, n_components should be <= number of attributes ({n})"
)));
}
@@ -374,21 +373,20 @@ mod tests {
let parameters = PCASearchParameters {
n_components: vec![2, 4],
use_correlation_matrix: vec![true, false],
..Default::default()
};
let mut iter = parameters.into_iter();
let next = iter.next().unwrap();
assert_eq!(next.n_components, 2);
assert_eq!(next.use_correlation_matrix, true);
assert!(next.use_correlation_matrix);
let next = iter.next().unwrap();
assert_eq!(next.n_components, 4);
assert_eq!(next.use_correlation_matrix, true);
assert!(next.use_correlation_matrix);
let next = iter.next().unwrap();
assert_eq!(next.n_components, 2);
assert_eq!(next.use_correlation_matrix, false);
assert!(!next.use_correlation_matrix);
let next = iter.next().unwrap();
assert_eq!(next.n_components, 4);
assert_eq!(next.use_correlation_matrix, false);
assert!(!next.use_correlation_matrix);
assert!(iter.next().is_none());
}
@@ -572,8 +570,8 @@ mod tests {
epsilon = 1e-4
));
for i in 0..pca.eigenvalues.len() {
assert!((pca.eigenvalues[i].abs() - expected_eigenvalues[i].abs()).abs() < 1e-8);
for (i, pca_eigenvalues_i) in pca.eigenvalues.iter().enumerate() {
assert!((pca_eigenvalues_i.abs() - expected_eigenvalues[i].abs()).abs() < 1e-8);
}
let us_arrests_t = pca.transform(&us_arrests).unwrap();
@@ -694,8 +692,8 @@ mod tests {
epsilon = 1e-4
));
for i in 0..pca.eigenvalues.len() {
assert!((pca.eigenvalues[i].abs() - expected_eigenvalues[i].abs()).abs() < 1e-8);
for (i, pca_eigenvalues_i) in pca.eigenvalues.iter().enumerate() {
assert!((pca_eigenvalues_i.abs() - expected_eigenvalues[i].abs()).abs() < 1e-8);
}
let us_arrests_t = pca.transform(&us_arrests).unwrap();
src/decomposition/svd.rs
+2 -5
View File
@@ -180,8 +180,7 @@ impl<T: Number + RealNumber, X: Array2<T> + SVDDecomposable<T> + EVDDecomposable
if parameters.n_components >= p {
return Err(Failed::fit(&format!(
"Number of components, n_components should be < number of attributes ({})",
p
"Number of components, n_components should be < number of attributes ({p})"
)));
}
@@ -202,8 +201,7 @@ impl<T: Number + RealNumber, X: Array2<T> + SVDDecomposable<T> + EVDDecomposable
let (p_c, k) = self.components.shape();
if p_c != p {
return Err(Failed::transform(&format!(
"Can not transform a {}x{} matrix into {}x{} matrix, incorrect input dimentions",
n, p, n, k
"Can not transform a {n}x{p} matrix into {n}x{k} matrix, incorrect input dimentions"
)));
}
@@ -227,7 +225,6 @@ mod tests {
fn search_parameters() {
let parameters = SVDSearchParameters {
n_components: vec![10, 100],
..Default::default()
};
let mut iter = parameters.into_iter();
let next = iter.next().unwrap();
src/error/mod.rs
+1 -1
View File
@@ -98,7 +98,7 @@ impl fmt::Display for FailedError {
FailedError::SolutionFailed => "Can't find solution",
FailedError::ParametersError => "Error in input, check parameters",
};
write!(f, "{}", failed_err_str)
write!(f, "{failed_err_str}")
}
}
src/lib.rs
+2 -1
View File
@@ -3,7 +3,8 @@
clippy::too_many_arguments,
clippy::many_single_char_names,
clippy::unnecessary_wraps,
clippy::upper_case_acronyms
clippy::upper_case_acronyms,
clippy::approx_constant
)]
#![warn(missing_docs)]
#![warn(rustdoc::missing_doc_code_examples)]
src/linalg/basic/arrays.rs
+12 -29
View File
@@ -548,7 +548,7 @@ pub trait ArrayView2<T: Debug + Display + Copy + Sized>: Array<T, (usize, usize)
let (nrows, ncols) = self.shape();
for r in 0..nrows {
let row: Vec<T> = (0..ncols).map(|c| *self.get((r, c))).collect();
writeln!(f, "{:?}", row)?
writeln!(f, "{row:?}")?
}
Ok(())
}
@@ -918,8 +918,7 @@ pub trait Array1<T: Debug + Display + Copy + Sized>: MutArrayView1<T> + Sized +
let len = self.shape();
assert!(
index.iter().all(|&i| i < len),
"All indices in `take` should be < {}",
len
"All indices in `take` should be < {len}"
);
Self::from_iterator(index.iter().map(move |&i| *self.get(i)), index.len())
}
@@ -990,10 +989,7 @@ pub trait Array1<T: Debug + Display + Copy + Sized>: MutArrayView1<T> + Sized +
};
assert!(
d1 == len,
"Can not multiply {}x{} matrix by {} vector",
nrows,
ncols,
len
"Can not multiply {nrows}x{ncols} matrix by {len} vector"
);
let mut result = Self::zeros(d2);
for i in 0..d2 {
@@ -1111,11 +1107,7 @@ pub trait Array2<T: Debug + Display + Copy + Sized>: MutArrayView2<T> + Sized +
assert!(
nrows * ncols == onrows * oncols,
"Can't reshape {}x{} array into a {}x{} array",
onrows,
oncols,
nrows,
ncols
"Can't reshape {onrows}x{oncols} array into a {nrows}x{ncols} array"
);
Self::from_iterator(self.iterator(0).cloned(), nrows, ncols, axis)
@@ -1129,11 +1121,7 @@ pub trait Array2<T: Debug + Display + Copy + Sized>: MutArrayView2<T> + Sized +
let (o_nrows, o_ncols) = other.shape();
assert!(
ncols == o_nrows,
"Can't multiply {}x{} and {}x{} matrices",
nrows,
ncols,
o_nrows,
o_ncols
"Can't multiply {nrows}x{ncols} and {o_nrows}x{o_ncols} matrices"
);
let inner_d = ncols;
let mut result = Self::zeros(nrows, o_ncols);
@@ -1166,7 +1154,7 @@ pub trait Array2<T: Debug + Display + Copy + Sized>: MutArrayView2<T> + Sized +
_ => (nrows, ncols, o_nrows, o_ncols),
};
if d1 != d4 {
panic!("Can not multiply {}x{} by {}x{} matrices", d2, d1, d4, d3);
panic!("Can not multiply {d2}x{d1} by {d4}x{d3} matrices");
}
let mut result = Self::zeros(d2, d3);
for r in 0..d2 {
@@ -1198,10 +1186,7 @@ pub trait Array2<T: Debug + Display + Copy + Sized>: MutArrayView2<T> + Sized +
};
assert!(
d2 == len,
"Can not multiply {}x{} matrix by {} vector",
nrows,
ncols,
len
"Can not multiply {nrows}x{ncols} matrix by {len} vector"
);
let mut result = Self::zeros(d1, 1);
for i in 0..d1 {
@@ -1432,8 +1417,7 @@ pub trait Array2<T: Debug + Display + Copy + Sized>: MutArrayView2<T> + Sized +
0 => {
assert!(
index.iter().all(|&i| i < nrows),
"All indices in `take` should be < {}",
nrows
"All indices in `take` should be < {nrows}"
);
Self::from_iterator(
index
@@ -1448,8 +1432,7 @@ pub trait Array2<T: Debug + Display + Copy + Sized>: MutArrayView2<T> + Sized +
_ => {
assert!(
index.iter().all(|&i| i < ncols),
"All indices in `take` should be < {}",
ncols
"All indices in `take` should be < {ncols}"
);
Self::from_iterator(
(0..nrows)
@@ -1736,7 +1719,7 @@ mod tests {
let r = Vec::<f32>::rand(4);
assert!(r.iterator(0).all(|&e| e <= 1f32));
assert!(r.iterator(0).all(|&e| e >= 0f32));
assert!(r.iterator(0).map(|v| *v).sum::<f32>() > 0f32);
assert!(r.iterator(0).copied().sum::<f32>() > 0f32);
}
#[test]
@@ -1954,7 +1937,7 @@ mod tests {
DenseMatrix::from_2d_array(&[&[1, 3], &[2, 4]])
);
assert_eq!(
DenseMatrix::concatenate_2d(&[&a.clone(), &b.clone()], 0),
DenseMatrix::concatenate_2d(&[&a, &b], 0),
DenseMatrix::from_2d_array(&[&[1, 2], &[3, 4], &[5, 6], &[7, 8]])
);
assert_eq!(
@@ -2025,7 +2008,7 @@ mod tests {
let r = DenseMatrix::<f32>::rand(2, 2);
assert!(r.iterator(0).all(|&e| e <= 1f32));
assert!(r.iterator(0).all(|&e| e >= 0f32));
assert!(r.iterator(0).map(|v| *v).sum::<f32>() > 0f32);
assert!(r.iterator(0).copied().sum::<f32>() > 0f32);
}
#[test]
src/linalg/basic/matrix.rs
+9 -9
View File
@@ -581,9 +581,9 @@ mod tests {
vec![4, 5, 6],
DenseMatrix::from_slice(&(*x.slice(1..2, 0..3))).values
);
let second_row: Vec<i32> = x.slice(1..2, 0..3).iterator(0).map(|x| *x).collect();
let second_row: Vec<i32> = x.slice(1..2, 0..3).iterator(0).copied().collect();
assert_eq!(vec![4, 5, 6], second_row);
let second_col: Vec<i32> = x.slice(0..3, 1..2).iterator(0).map(|x| *x).collect();
let second_col: Vec<i32> = x.slice(0..3, 1..2).iterator(0).copied().collect();
assert_eq!(vec![2, 5, 8], second_col);
}
@@ -640,12 +640,12 @@ mod tests {
let x = DenseMatrix::<&str>::from_2d_array(&[&["1", "2", "3"], &["4", "5", "6"]]);
assert_eq!(vec!["1", "4", "2", "5", "3", "6"], x.values);
assert!(x.column_major == true);
assert!(x.column_major);
// transpose
let x = x.transpose();
assert_eq!(vec!["1", "4", "2", "5", "3", "6"], x.values);
assert!(x.column_major == false); // should change column_major
assert!(!x.column_major); // should change column_major
}
#[test]
@@ -659,7 +659,7 @@ mod tests {
vec![1, 2, 3, 4, 5, 6],
m.values.iter().map(|e| **e).collect::<Vec<i32>>()
);
assert!(m.column_major == false);
assert!(!m.column_major);
}
#[test]
@@ -667,10 +667,10 @@ mod tests {
let a = DenseMatrix::from_2d_array(&[&[1, 2, 3], &[4, 5, 6]]);
let b = DenseMatrix::from_2d_array(&[&[1, 2], &[3, 4], &[5, 6]]);
println!("{}", a);
println!("{a}");
// take column 0 and 2
assert_eq!(vec![1, 3, 4, 6], a.take(&[0, 2], 1).values);
println!("{}", b);
println!("{b}");
// take rows 0 and 2
assert_eq!(vec![1, 2, 5, 6], b.take(&[0, 2], 0).values);
}
@@ -692,11 +692,11 @@ mod tests {
let a = a.reshape(2, 6, 0);
assert_eq!(vec![1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12], a.values);
assert!(a.ncols == 6 && a.nrows == 2 && a.column_major == false);
assert!(a.ncols == 6 && a.nrows == 2 && !a.column_major);
let a = a.reshape(3, 4, 1);
assert_eq!(vec![1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12], a.values);
assert!(a.ncols == 4 && a.nrows == 3 && a.column_major == true);
assert!(a.ncols == 4 && a.nrows == 3 && a.column_major);
}
#[test]
src/linalg/basic/vector.rs
+3 -3
View File
@@ -160,8 +160,8 @@ mod tests {
fn dot_product<T: Number, V: Array1<T>>(v: &V) -> T {
let vv = V::zeros(10);
let v_s = vv.slice(0..3);
let dot = v_s.dot(v);
dot
v_s.dot(v)
}
fn vector_ops<T: Number + PartialOrd, V: Array1<T>>(_: &V) -> T {
@@ -216,7 +216,7 @@ mod tests {
#[test]
fn test_mut_iterator() {
let mut x = vec![1, 2, 3];
x.iterator_mut(0).for_each(|v| *v = *v * 2);
x.iterator_mut(0).for_each(|v| *v *= 2);
assert_eq!(vec![2, 4, 6], x);
}
src/linalg/ndarray/matrix.rs
+6 -6
View File
@@ -217,7 +217,7 @@ mod tests {
fn test_iterator() {
let a = arr2(&[[1, 2, 3], [4, 5, 6]]);
let v: Vec<i32> = a.iterator(0).map(|&v| v).collect();
let v: Vec<i32> = a.iterator(0).copied().collect();
assert_eq!(v, vec!(1, 2, 3, 4, 5, 6));
}
@@ -236,7 +236,7 @@ mod tests {
let x = arr2(&[[1, 2, 3], [4, 5, 6]]);
let x_slice = Array2::slice(&x, 0..2, 1..2);
assert_eq!((2, 1), x_slice.shape());
let v: Vec<i32> = x_slice.iterator(0).map(|&v| v).collect();
let v: Vec<i32> = x_slice.iterator(0).copied().collect();
assert_eq!(v, [2, 5]);
}
@@ -245,11 +245,11 @@ mod tests {
let x = arr2(&[[1, 2, 3], [4, 5, 6]]);
let x_slice = Array2::slice(&x, 0..2, 0..3);
assert_eq!(
x_slice.iterator(0).map(|&v| v).collect::<Vec<i32>>(),
x_slice.iterator(0).copied().collect::<Vec<i32>>(),
vec![1, 2, 3, 4, 5, 6]
);
assert_eq!(
x_slice.iterator(1).map(|&v| v).collect::<Vec<i32>>(),
x_slice.iterator(1).copied().collect::<Vec<i32>>(),
vec![1, 4, 2, 5, 3, 6]
);
}
@@ -279,8 +279,8 @@ mod tests {
fn test_c_from_iterator() {
let data = vec![1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12];
let a: NDArray2<i32> = Array2::from_iterator(data.clone().into_iter(), 4, 3, 0);
println!("{}", a);
println!("{a}");
let a: NDArray2<i32> = Array2::from_iterator(data.into_iter(), 4, 3, 1);
println!("{}", a);
println!("{a}");
}
}
src/linalg/ndarray/vector.rs
+1 -1
View File
@@ -152,7 +152,7 @@ mod tests {
fn test_iterator() {
let a = arr1(&[1, 2, 3]);
let v: Vec<i32> = a.iterator(0).map(|&v| v).collect();
let v: Vec<i32> = a.iterator(0).copied().collect();
assert_eq!(v, vec!(1, 2, 3));
}
src/linalg/traits/evd.rs
+9 -13
View File
@@ -66,7 +66,7 @@ pub trait EVDDecomposable<T: Number + RealNumber>: Array2<T> {
fn evd_mut(mut self, symmetric: bool) -> Result<EVD<T, Self>, Failed> {
let (nrows, ncols) = self.shape();
if ncols != nrows {
panic!("Matrix is not square: {} x {}", nrows, ncols);
panic!("Matrix is not square: {nrows} x {ncols}");
}
let n = nrows;
@@ -837,10 +837,8 @@ mod tests {
evd.V.abs(),
epsilon = 1e-4
));
for i in 0..eigen_values.len() {
assert!((eigen_values[i] - evd.d[i]).abs() < 1e-4);
}
for i in 0..eigen_values.len() {
for (i, eigen_values_i) in eigen_values.iter().enumerate() {
assert!((eigen_values_i - evd.d[i]).abs() < 1e-4);
assert!((0f64 - evd.e[i]).abs() < std::f64::EPSILON);
}
}
@@ -871,10 +869,8 @@ mod tests {
evd.V.abs(),
epsilon = 1e-4
));
for i in 0..eigen_values.len() {
assert!((eigen_values[i] - evd.d[i]).abs() < 1e-4);
}
for i in 0..eigen_values.len() {
for (i, eigen_values_i) in eigen_values.iter().enumerate() {
assert!((eigen_values_i - evd.d[i]).abs() < 1e-4);
assert!((0f64 - evd.e[i]).abs() < std::f64::EPSILON);
}
}
@@ -908,11 +904,11 @@ mod tests {
evd.V.abs(),
epsilon = 1e-4
));
for i in 0..eigen_values_d.len() {
assert!((eigen_values_d[i] - evd.d[i]).abs() < 1e-4);
for (i, eigen_values_d_i) in eigen_values_d.iter().enumerate() {
assert!((eigen_values_d_i - evd.d[i]).abs() < 1e-4);
}
for i in 0..eigen_values_e.len() {
assert!((eigen_values_e[i] - evd.e[i]).abs() < 1e-4);
for (i, eigen_values_e_i) in eigen_values_e.iter().enumerate() {
assert!((eigen_values_e_i - evd.e[i]).abs() < 1e-4);
}
}
}
src/linalg/traits/lu.rs
+2 -5
View File
@@ -126,7 +126,7 @@ impl<T: Number + RealNumber, M: Array2<T>> LU<T, M> {
let (m, n) = self.LU.shape();
if m != n {
panic!("Matrix is not square: {}x{}", m, n);
panic!("Matrix is not square: {m}x{n}");
}
let mut inv = M::zeros(n, n);
@@ -143,10 +143,7 @@ impl<T: Number + RealNumber, M: Array2<T>> LU<T, M> {
let (b_m, b_n) = b.shape();
if b_m != m {
panic!(
"Row dimensions do not agree: A is {} x {}, but B is {} x {}",
m, n, b_m, b_n
);
panic!("Row dimensions do not agree: A is {m} x {n}, but B is {b_m} x {b_n}");
}
if self.singular {
src/linalg/traits/qr.rs
+1 -4
View File
@@ -102,10 +102,7 @@ impl<T: Number + RealNumber, M: Array2<T>> QR<T, M> {
let (b_nrows, b_ncols) = b.shape();
if b_nrows != m {
panic!(
"Row dimensions do not agree: A is {} x {}, but B is {} x {}",
m, n, b_nrows, b_ncols
);
panic!("Row dimensions do not agree: A is {m} x {n}, but B is {b_nrows} x {b_ncols}");
}
if self.singular {
src/linalg/traits/stats.rs
+1 -1
View File
@@ -286,7 +286,7 @@ mod tests {
}
{
let mut m = m.clone();
let mut m = m;
m.standard_scale_mut(&m.mean(1), &m.std(1), 1);
assert_eq!(&m, &expected_1);
}
src/linalg/traits/svd.rs
+4 -4
View File
@@ -509,8 +509,8 @@ mod tests {
assert!(relative_eq!(V.abs(), svd.V.abs(), epsilon = 1e-4));
assert!(relative_eq!(U.abs(), svd.U.abs(), epsilon = 1e-4));
for i in 0..s.len() {
assert!((s[i] - svd.s[i]).abs() < 1e-4);
for (i, s_i) in s.iter().enumerate() {
assert!((s_i - svd.s[i]).abs() < 1e-4);
}
}
#[cfg_attr(
@@ -713,8 +713,8 @@ mod tests {
assert!(relative_eq!(V.abs(), svd.V.abs(), epsilon = 1e-4));
assert!(relative_eq!(U.abs(), svd.U.abs(), epsilon = 1e-4));
for i in 0..s.len() {
assert!((s[i] - svd.s[i]).abs() < 1e-4);
for (i, s_i) in s.iter().enumerate() {
assert!((s_i - svd.s[i]).abs() < 1e-4);
}
}
#[cfg_attr(
src/linear/elastic_net.rs
+1 -4
View File
@@ -425,10 +425,7 @@ impl<TX: FloatNumber + RealNumber, TY: Number, X: Array2<TX>, Y: Array1<TY>>
for (i, col_std_i) in col_std.iter().enumerate() {
if (*col_std_i - TX::zero()).abs() < TX::epsilon() {
return Err(Failed::fit(&format!(
"Cannot rescale constant column {}",
i
)));
return Err(Failed::fit(&format!("Cannot rescale constant column {i}")));
}
}
src/linear/lasso.rs
+1 -4
View File
@@ -356,10 +356,7 @@ impl<TX: FloatNumber + RealNumber, TY: Number, X: Array2<TX>, Y: Array1<TY>> Las
for (i, col_std_i) in col_std.iter().enumerate() {
if (*col_std_i - TX::zero()).abs() < TX::epsilon() {
return Err(Failed::fit(&format!(
"Cannot rescale constant column {}",
i
)));
return Err(Failed::fit(&format!("Cannot rescale constant column {i}")));
}
}
src/linear/logistic_regression.rs
+7 -8
View File
@@ -449,8 +449,7 @@ impl<TX: Number + FloatNumber + RealNumber, TY: Number + Ord, X: Array2<TX>, Y:
match k.cmp(&2) {
Ordering::Less => Err(Failed::fit(&format!(
"incorrect number of classes: {}. Should be >= 2.",
k
"incorrect number of classes: {k}. Should be >= 2."
))),
Ordering::Equal => {
let x0 = Vec::zeros(num_attributes + 1);
@@ -636,19 +635,19 @@ mod tests {
assert!((g[0] + 33.000068218163484).abs() < std::f64::EPSILON);
let f = objective.f(&vec![1., 2., 3., 4., 5., 6., 7., 8., 9.]);
let f = objective.f(&[1., 2., 3., 4., 5., 6., 7., 8., 9.]);
assert!((f - 408.0052230582765).abs() < std::f64::EPSILON);
let objective_reg = MultiClassObjectiveFunction {
x: &x,
y: y.clone(),
y,
k: 3,
alpha: 1.0,
_phantom_t: PhantomData,
};
let f = objective_reg.f(&vec![1., 2., 3., 4., 5., 6., 7., 8., 9.]);
let f = objective_reg.f(&[1., 2., 3., 4., 5., 6., 7., 8., 9.]);
assert!((f - 487.5052).abs() < 1e-4);
objective_reg.df(&mut g, &vec![1., 2., 3., 4., 5., 6., 7., 8., 9.]);
@@ -697,18 +696,18 @@ mod tests {
assert!((g[1] - 10.239000702928523).abs() < std::f64::EPSILON);
assert!((g[2] - 3.869294270156324).abs() < std::f64::EPSILON);
let f = objective.f(&vec![1., 2., 3.]);
let f = objective.f(&[1., 2., 3.]);
assert!((f - 59.76994756647412).abs() < std::f64::EPSILON);
let objective_reg = BinaryObjectiveFunction {
x: &x,
y: y.clone(),
y,
alpha: 1.0,
_phantom_t: PhantomData,
};
let f = objective_reg.f(&vec![1., 2., 3.]);
let f = objective_reg.f(&[1., 2., 3.]);
assert!((f - 62.2699).abs() < 1e-4);
objective_reg.df(&mut g, &vec![1., 2., 3.]);
src/linear/ridge_regression.rs
+1 -4
View File
@@ -384,10 +384,7 @@ impl<
for (i, col_std_i) in col_std.iter().enumerate() {
if (*col_std_i - TX::zero()).abs() < TX::epsilon() {
return Err(Failed::fit(&format!(
"Cannot rescale constant column {}",
i
)));
return Err(Failed::fit(&format!("Cannot rescale constant column {i}")));
}
}
src/metrics/cluster_hcv.rs
+3 -3
View File
@@ -98,8 +98,8 @@ mod tests {
let mut scores = HCVScore::new();
scores.compute(&v1, &v2);
assert!((0.2548 - scores.homogeneity.unwrap() as f64).abs() < 1e-4);
assert!((0.5440 - scores.completeness.unwrap() as f64).abs() < 1e-4);
assert!((0.3471 - scores.v_measure.unwrap() as f64).abs() < 1e-4);
assert!((0.2548 - scores.homogeneity.unwrap()).abs() < 1e-4);
assert!((0.5440 - scores.completeness.unwrap()).abs() < 1e-4);
assert!((0.3471 - scores.v_measure.unwrap()).abs() < 1e-4);
}
}
src/metrics/cluster_helpers.rs
+1 -1
View File
@@ -125,7 +125,7 @@ mod tests {
fn entropy_test() {
let v1 = vec![0, 0, 1, 1, 2, 0, 4];
assert!((1.2770 - entropy(&v1).unwrap() as f64).abs() < 1e-4);
assert!((1.2770 - entropy(&v1).unwrap()).abs() < 1e-4);
}
#[cfg_attr(
src/metrics/f1.rs
+2 -2
View File
@@ -95,8 +95,8 @@ mod tests {
let score1: f64 = F1::new_with(beta).get_score(&y_true, &y_pred);
let score2: f64 = F1::new_with(beta).get_score(&y_true, &y_true);
println!("{:?}", score1);
println!("{:?}", score2);
println!("{score1:?}");
println!("{score2:?}");
assert!((score1 - 0.57142857).abs() < 1e-8);
assert!((score2 - 1.0).abs() < 1e-8);
src/model_selection/kfold.rs
+4 -4
View File
@@ -213,17 +213,17 @@ mod tests {
for t in &test_masks[0][0..11] {
// TODO: this can be prob done better
assert_eq!(*t, true)
assert!(*t)
}
for t in &test_masks[0][11..22] {
assert_eq!(*t, false)
assert!(!*t)
}
for t in &test_masks[1][0..11] {
assert_eq!(*t, false)
assert!(!*t)
}
for t in &test_masks[1][11..22] {
assert_eq!(*t, true)
assert!(*t)
}
}
src/model_selection/mod.rs
+2 -2
View File
@@ -169,7 +169,7 @@ pub fn train_test_split<
let n_test = ((n as f32) * test_size) as usize;
if n_test < 1 {
panic!("number of sample is too small {}", n);
panic!("number of sample is too small {n}");
}
let mut indices: Vec<usize> = (0..n).collect();
@@ -553,6 +553,6 @@ mod tests {
&accuracy,
)
.unwrap();
println!("{:?}", results);
println!("{results:?}");
}
}
src/naive_bayes/bernoulli.rs
+4 -8
View File
@@ -271,21 +271,18 @@ impl<TY: Number + Ord + Unsigned> BernoulliNBDistribution<TY> {
let y_samples = y.shape();
if y_samples != n_samples {
return Err(Failed::fit(&format!(
"Size of x should equal size of y; |x|=[{}], |y|=[{}]",
n_samples, y_samples
"Size of x should equal size of y; |x|=[{n_samples}], |y|=[{y_samples}]"
)));
}
if n_samples == 0 {
return Err(Failed::fit(&format!(
"Size of x and y should greater than 0; |x|=[{}]",
n_samples
"Size of x and y should greater than 0; |x|=[{n_samples}]"
)));
}
if alpha < 0f64 {
return Err(Failed::fit(&format!(
"Alpha should be greater than 0; |alpha|=[{}]",
alpha
"Alpha should be greater than 0; |alpha|=[{alpha}]"
)));
}
@@ -318,8 +315,7 @@ impl<TY: Number + Ord + Unsigned> BernoulliNBDistribution<TY> {
feature_in_class_counter[class_index][idx] +=
row_i.to_usize().ok_or_else(|| {
Failed::fit(&format!(
"Elements of the matrix should be 1.0 or 0.0 |found|=[{}]",
row_i
"Elements of the matrix should be 1.0 or 0.0 |found|=[{row_i}]"
))
})?;
}
src/naive_bayes/categorical.rs
+4 -9
View File
@@ -158,8 +158,7 @@ impl<T: Number + Unsigned> CategoricalNBDistribution<T> {
pub fn fit<X: Array2<T>, Y: Array1<T>>(x: &X, y: &Y, alpha: f64) -> Result<Self, Failed> {
if alpha < 0f64 {
return Err(Failed::fit(&format!(
"alpha should be >= 0, alpha=[{}]",
alpha
"alpha should be >= 0, alpha=[{alpha}]"
)));
}
@@ -167,15 +166,13 @@ impl<T: Number + Unsigned> CategoricalNBDistribution<T> {
let y_samples = y.shape();
if y_samples != n_samples {
return Err(Failed::fit(&format!(
"Size of x should equal size of y; |x|=[{}], |y|=[{}]",
n_samples, y_samples
"Size of x should equal size of y; |x|=[{n_samples}], |y|=[{y_samples}]"
)));
}
if n_samples == 0 {
return Err(Failed::fit(&format!(
"Size of x and y should greater than 0; |x|=[{}]",
n_samples
"Size of x and y should greater than 0; |x|=[{n_samples}]"
)));
}
let y: Vec<usize> = y.iterator(0).map(|y_i| y_i.to_usize().unwrap()).collect();
@@ -202,8 +199,7 @@ impl<T: Number + Unsigned> CategoricalNBDistribution<T> {
.max()
.ok_or_else(|| {
Failed::fit(&format!(
"Failed to get the categories for feature = {}",
feature
"Failed to get the categories for feature = {feature}"
))
})?;
n_categories.push(feature_max + 1);
@@ -429,7 +425,6 @@ mod tests {
fn search_parameters() {
let parameters = CategoricalNBSearchParameters {
alpha: vec![1., 2.],
..Default::default()
};
let mut iter = parameters.into_iter();
let next = iter.next().unwrap();
src/naive_bayes/gaussian.rs
+2 -5
View File
@@ -185,15 +185,13 @@ impl<TY: Number + Ord + Unsigned> GaussianNBDistribution<TY> {
let y_samples = y.shape();
if y_samples != n_samples {
return Err(Failed::fit(&format!(
"Size of x should equal size of y; |x|=[{}], |y|=[{}]",
n_samples, y_samples
"Size of x should equal size of y; |x|=[{n_samples}], |y|=[{y_samples}]"
)));
}
if n_samples == 0 {
return Err(Failed::fit(&format!(
"Size of x and y should greater than 0; |x|=[{}]",
n_samples
"Size of x and y should greater than 0; |x|=[{n_samples}]"
)));
}
let (class_labels, indices) = y.unique_with_indices();
@@ -375,7 +373,6 @@ mod tests {
fn search_parameters() {
let parameters = GaussianNBSearchParameters {
priors: vec![Some(vec![1.]), Some(vec![2.])],
..Default::default()
};
let mut iter = parameters.into_iter();
let next = iter.next().unwrap();
src/naive_bayes/multinomial.rs
+4 -8
View File
@@ -220,21 +220,18 @@ impl<TY: Number + Ord + Unsigned> MultinomialNBDistribution<TY> {
let y_samples = y.shape();
if y_samples != n_samples {
return Err(Failed::fit(&format!(
"Size of x should equal size of y; |x|=[{}], |y|=[{}]",
n_samples, y_samples
"Size of x should equal size of y; |x|=[{n_samples}], |y|=[{y_samples}]"
)));
}
if n_samples == 0 {
return Err(Failed::fit(&format!(
"Size of x and y should greater than 0; |x|=[{}]",
n_samples
"Size of x and y should greater than 0; |x|=[{n_samples}]"
)));
}
if alpha < 0f64 {
return Err(Failed::fit(&format!(
"Alpha should be greater than 0; |alpha|=[{}]",
alpha
"Alpha should be greater than 0; |alpha|=[{alpha}]"
)));
}
@@ -266,8 +263,7 @@ impl<TY: Number + Ord + Unsigned> MultinomialNBDistribution<TY> {
feature_in_class_counter[class_index][idx] +=
row_i.to_usize().ok_or_else(|| {
Failed::fit(&format!(
"Elements of the matrix should be convertible to usize |found|=[{}]",
row_i
"Elements of the matrix should be convertible to usize |found|=[{row_i}]"
))
})?;
}
src/neighbors/knn_classifier.rs
+1 -2
View File
@@ -236,8 +236,7 @@ impl<TX: Number, TY: Number + Ord, X: Array2<TX>, Y: Array1<TY>, D: Distance<Vec
if x_n != y_n {
return Err(Failed::fit(&format!(
"Size of x should equal size of y; |x|=[{}], |y|=[{}]",
x_n, y_n
"Size of x should equal size of y; |x|=[{x_n}], |y|=[{y_n}]"
)));
}
src/neighbors/knn_regressor.rs
+1 -2
View File
@@ -224,8 +224,7 @@ impl<TX: Number, TY: Number, X: Array2<TX>, Y: Array1<TY>, D: Distance<Vec<TX>>>
if x_n != y_n {
return Err(Failed::fit(&format!(
"Size of x should equal size of y; |x|=[{}], |y|=[{}]",
x_n, y_n
"Size of x should equal size of y; |x|=[{x_n}], |y|=[{y_n}]"
)));
}
src/optimization/first_order/gradient_descent.rs
+4 -3
View File
@@ -113,12 +113,13 @@ mod tests {
g[1] = 200. * (x[1] - x[0].powf(2.));
};
let mut ls: Backtracking<f64> = Default::default();
ls.order = FunctionOrder::THIRD;
let ls: Backtracking<f64> = Backtracking::<f64> {
order: FunctionOrder::THIRD,
..Default::default()
};
let optimizer: GradientDescent = Default::default();
let result = optimizer.optimize(&f, &df, &x0, &ls);
println!("{:?}", result);
assert!((result.f_x - 0.0).abs() < 1e-5);
assert!((result.x[0] - 1.0).abs() < 1e-2);
src/optimization/first_order/lbfgs.rs
+6 -4
View File
@@ -196,9 +196,9 @@ impl LBFGS {
}
///
fn update_hessian<'a, T: FloatNumber, X: Array1<T>>(
fn update_hessian<T: FloatNumber, X: Array1<T>>(
&self,
_: &'a DF<'_, X>,
_: &DF<'_, X>,
state: &mut LBFGSState<T, X>,
) {
state.dg = state.x_df.sub(&state.x_df_prev);
@@ -291,8 +291,10 @@ mod tests {
g[0] = -2. * (1. - x[0]) - 400. * (x[1] - x[0].powf(2.)) * x[0];
g[1] = 200. * (x[1] - x[0].powf(2.));
};
let mut ls: Backtracking<f64> = Default::default();
ls.order = FunctionOrder::THIRD;
let ls: Backtracking<f64> = Backtracking::<f64> {
order: FunctionOrder::THIRD,
..Default::default()
};
let optimizer: LBFGS = Default::default();
let result = optimizer.optimize(&f, &df, &x0, &ls);
src/preprocessing/categorical.rs
+4 -9
View File
@@ -132,8 +132,7 @@ impl OneHotEncoder {
data.copy_col_as_vec(idx, &mut col_buf);
if !validate_col_is_categorical(&col_buf) {
let msg = format!(
"Column {} of data matrix containts non categorizable (integer) values",
idx
"Column {idx} of data matrix containts non categorizable (integer) values"
);
return Err(Failed::fit(&msg[..]));
}
@@ -182,7 +181,7 @@ impl OneHotEncoder {
match oh_vec {
None => {
// Since we support T types, bad value in a series causes in to be invalid
let msg = format!("At least one value in column {} doesn't conform to category definition", old_cidx);
let msg = format!("At least one value in column {old_cidx} doesn't conform to category definition");
return Err(Failed::transform(&msg[..]));
}
Some(v) => {
@@ -338,11 +337,7 @@ mod tests {
]);
let params = OneHotEncoderParams::from_cat_idx(&[1]);
match OneHotEncoder::fit(&m, params) {
Err(_) => {
assert!(true);
}
_ => assert!(false),
}
let result = OneHotEncoder::fit(&m, params);
assert!(result.is_err());
}
}
src/preprocessing/numerical.rs
+1 -1
View File
@@ -294,7 +294,7 @@ mod tests {
&[0.5708488802, 0.1846414616, 0.9590802982, 0.5591871046],
&[0.8387612750, 0.5754861361, 0.5537109852, 0.1077646442],
]));
println!("{}", transformed_values);
println!("{transformed_values}");
assert!(transformed_values.approximate_eq(
&DenseMatrix::from_2d_array(&[
&[-1.1154020653, -0.4031985330, 0.9284605204, -0.4271473866],
src/preprocessing/series_encoder.rs
+4 -4
View File
@@ -206,7 +206,7 @@ mod tests {
#[test]
fn from_categories() {
let fake_categories: Vec<usize> = vec![1, 2, 3, 4, 5, 3, 5, 3, 1, 2, 4];
let it = fake_categories.iter().map(|&a| a);
let it = fake_categories.iter().copied();
let enc = CategoryMapper::<usize>::fit_to_iter(it);
let oh_vec: Vec<f64> = match enc.get_one_hot(&1) {
None => panic!("Wrong categories"),
@@ -218,8 +218,8 @@ mod tests {
fn build_fake_str_enc<'a>() -> CategoryMapper<&'a str> {
let fake_category_pos = vec!["background", "dog", "cat"];
let enc = CategoryMapper::<&str>::from_positional_category_vec(fake_category_pos);
enc
CategoryMapper::<&str>::from_positional_category_vec(fake_category_pos)
}
#[cfg_attr(
all(target_arch = "wasm32", not(target_os = "wasi")),
@@ -275,7 +275,7 @@ mod tests {
let lab = enc.invert_one_hot(res).unwrap();
assert_eq!(lab, "dog");
if let Err(e) = enc.invert_one_hot(vec![0.0, 0.0, 0.0]) {
let pos_entries = format!("Expected a single positive entry, 0 entires found");
let pos_entries = "Expected a single positive entry, 0 entires found".to_string();
assert_eq!(e, Failed::transform(&pos_entries[..]));
};
}
src/readers/csv.rs
+2 -2
View File
@@ -167,7 +167,7 @@ where
}
/// Ensure that a string containing a csv row conforms to a specified row format.
fn validate_csv_row<'a>(row: &'a str, row_format: &CSVRowFormat<'_>) -> Result<(), ReadingError> {
fn validate_csv_row(row: &str, row_format: &CSVRowFormat<'_>) -> Result<(), ReadingError> {
let actual_number_of_fields = row.split(row_format.field_seperator).count();
if row_format.n_fields == actual_number_of_fields {
Ok(())
@@ -208,7 +208,7 @@ where
match value_string.parse::<T>().ok() {
Some(value) => Ok(value),
None => Err(ReadingError::InvalidField {
msg: format!("Value '{}' could not be read.", value_string,),
msg: format!("Value '{value_string}' could not be read.",),
}),
}
}
src/svm/svc.rs
+2 -10
View File
@@ -983,11 +983,7 @@ mod tests {
.unwrap();
let acc = accuracy(&y, &(y_hat.iter().map(|e| e.to_i32().unwrap()).collect()));
assert!(
acc >= 0.9,
"accuracy ({}) is not larger or equal to 0.9",
acc
);
assert!(acc >= 0.9, "accuracy ({acc}) is not larger or equal to 0.9");
}
#[cfg_attr(
@@ -1076,11 +1072,7 @@ mod tests {
let acc = accuracy(&y, &(y_hat.iter().map(|e| e.to_i32().unwrap()).collect()));
assert!(
acc >= 0.9,
"accuracy ({}) is not larger or equal to 0.9",
acc
);
assert!(acc >= 0.9, "accuracy ({acc}) is not larger or equal to 0.9");
}
#[cfg_attr(
src/svm/svr.rs
+1 -1
View File
@@ -662,7 +662,7 @@ mod tests {
.unwrap();
let t = mean_squared_error(&y_hat, &y);
println!("{:?}", t);
println!("{t:?}");
assert!(t < 2.5);
}
src/tree/decision_tree_classifier.rs
+4 -7
View File
@@ -560,8 +560,7 @@ impl<TX: Number + PartialOrd, TY: Number + Ord, X: Array2<TX>, Y: Array1<TY>>
let k = classes.len();
if k < 2 {
return Err(Failed::fit(&format!(
"Incorrect number of classes: {}. Should be >= 2.",
k
"Incorrect number of classes: {k}. Should be >= 2."
)));
}
@@ -901,15 +900,13 @@ mod tests {
)]
#[test]
fn gini_impurity() {
assert!((impurity(&SplitCriterion::Gini, &[7, 3], 10) - 0.42).abs() < std::f64::EPSILON);
assert!(
(impurity(&SplitCriterion::Gini, &vec![7, 3], 10) - 0.42).abs() < std::f64::EPSILON
);
assert!(
(impurity(&SplitCriterion::Entropy, &vec![7, 3], 10) - 0.8812908992306927).abs()
(impurity(&SplitCriterion::Entropy, &[7, 3], 10) - 0.8812908992306927).abs()
< std::f64::EPSILON
);
assert!(
(impurity(&SplitCriterion::ClassificationError, &vec![7, 3], 10) - 0.3).abs()
(impurity(&SplitCriterion::ClassificationError, &[7, 3], 10) - 0.3).abs()
< std::f64::EPSILON
);
}