CI maintenance

.github/workflows/checking.yml

@@ -10,7 +10,7 @@ jobs:
       fail-fast: false
       matrix:
         toolchain:
-          - 1.70.0
+          - 1.71.1
           - stable
           - nightly
         os:

.github/workflows/codequality.yml

@@ -10,7 +10,7 @@ jobs:
     strategy:
       matrix:
         toolchain:
-          - 1.70.0
+          - 1.71.1
           - stable
 
     steps:
@@ -46,7 +46,7 @@ jobs:
         run: echo "::set-output name=version::$(cargo --version | cut -d ' ' -f 2)"
         shell: bash
 
-      - uses: actions/cache@v2
+      - uses: actions/cache@v4
         id: tarpaulin-cache
         with:
           path: |
@@ -61,6 +61,8 @@ jobs:
         run: |
           cargo tarpaulin --verbose --timeout 120 --out Xml --all --release
       - name: Upload to codecov.io
-        uses: codecov/codecov-action@v1
+        uses: codecov/codecov-action@v4
         with:
+          token: ${{ secrets.CODECOV_TOKEN }}
           fail_ci_if_error: true
+          verbose: true

.github/workflows/testing.yml

@@ -10,7 +10,7 @@ jobs:
       fail-fast: false
       matrix:
         toolchain:
-          - 1.70.0
+          - 1.71.1
           - stable
         os:
           - ubuntu-latest
@@ -35,7 +35,7 @@ jobs:
       fail-fast: false
       matrix:
         toolchain:
-          - 1.70.0
+          - 1.71.1
           - stable
         os:
           - ubuntu-latest

algorithms/linfa-bayes/README.md

@@ -10,8 +10,8 @@
 
 `linfa-bayes` currently provides an implementation of the following methods: 
 
-- Gaussian Naive Bayes ([`GaussianNb`](crate::GaussianNb))
-- Multinomial Naive Nayes ([`MultinomialNb`](crate::MultinomialNb))
+- Gaussian Naive Bayes ([`GaussianNb`])
+- Multinomial Naive Nayes ([`MultinomialNb`]))
 
 ## Examples
 

algorithms/linfa-bayes/src/base_nb.rs

@@ -79,7 +79,8 @@ pub fn filter<F: Float, L: Label + Ord>(
     let index = y
         .into_iter()
         .enumerate()
-        .filter_map(|(i, y)| (*ycondition == *y).then(|| i))
+        .filter(|(_, y)| (*ycondition == **y))
+        .map(|(i, _)| i)
         .collect::<Vec<_>>();
 
     // We subset x to only records corresponding to the class represented in `ycondition`

algorithms/linfa-bayes/src/gaussian_nb.rs

@@ -133,7 +133,7 @@ where
     }
 }
 
-impl<'a, F, L> GaussianNbValidParams<F, L>
+impl<F, L> GaussianNbValidParams<F, L>
 where
     F: Float,
 {
@@ -259,7 +259,7 @@ impl<F: Float, L: Label> GaussianNb<F, L> {
     }
 }
 
-impl<'a, F, L> NaiveBayes<'a, F, L> for GaussianNb<F, L>
+impl<F, L> NaiveBayes<'_, F, L> for GaussianNb<F, L>
 where
     F: Float,
     L: Label + Ord,

algorithms/linfa-bayes/src/multinomial_nb.rs

@@ -229,7 +229,7 @@ impl<F: Float, L: Label> MultinomialNb<F, L> {
     }
 }
 
-impl<'a, F, L> NaiveBayes<'a, F, L> for MultinomialNb<F, L>
+impl<F, L> NaiveBayes<'_, F, L> for MultinomialNb<F, L>
 where
     F: Float,
     L: Label + Ord,

algorithms/linfa-clustering/Cargo.toml

@@ -24,6 +24,7 @@ categories = ["algorithms", "mathematics", "science"]
 
 [features]
 default = []
+blas = []
 serde = ["serde_crate", "ndarray/serde", "linfa-nn/serde"]
 
 [dependencies.serde_crate]

algorithms/linfa-clustering/src/dbscan/algorithm.rs

@@ -38,10 +38,8 @@ use linfa::{traits::Transformer, DatasetBase};
 /// The algorithm iterates over each point in the dataset and for every point
 /// not yet assigned to a cluster:
 /// - Find all points within the neighborhood of size `tolerance`
-/// - If the number of points in the neighborhood is below a minimum size label
-/// as noise
-/// - Otherwise label the point with the cluster ID and repeat with each of the
-/// neighbours
+/// - If the number of points in the neighborhood is below a minimum size label as noise
+/// - Otherwise label the point with the cluster ID and repeat with each of the neighbours
 ///
 /// ## Tutorial
 ///

algorithms/linfa-clustering/src/gaussian_mixture/algorithm.rs

@@ -211,6 +211,7 @@ impl<F: Float> GaussianMixtureModel<F> {
         self.means()
     }
 
+    #[allow(clippy::type_complexity)]
     fn estimate_gaussian_parameters<D: Data<Elem = F>>(
         observations: &ArrayBase<D, Ix2>,
         resp: &Array2<F>,
@@ -505,17 +506,15 @@ mod tests {
     }
 
     pub struct MultivariateNormal {
-        pub mean: Array1<f64>,
-        pub covariance: Array2<f64>,
-        /// Lower triangular matrix (Cholesky decomposition of the coviariance matrix)
+        mean: Array1<f64>,
+        /// Lower triangular matrix (Cholesky decomposition of the covariance matrix)
         lower: Array2<f64>,
     }
     impl MultivariateNormal {
         pub fn new(mean: &ArrayView1<f64>, covariance: &ArrayView2<f64>) -> LAResult<Self> {
             let lower = covariance.cholesky()?;
             Ok(MultivariateNormal {
                 mean: mean.to_owned(),
-                covariance: covariance.to_owned(),
                 lower,
             })
         }

algorithms/linfa-clustering/src/k_means/algorithm.rs

@@ -766,7 +766,7 @@ mod tests {
             &mut rng,
         );
 
-        let expected_memberships = (0..n_centroids).into_iter().collect::<Array1<_>>();
+        let expected_memberships = (0..n_centroids).collect::<Array1<_>>();
         assert_eq!(
             calc_memberships!(L2Dist, centroids, centroids),
             expected_memberships

algorithms/linfa-clustering/src/k_means/init.rs

@@ -164,7 +164,7 @@ fn k_means_para<R: Rng, F: Float, D: Distance<F>>(
         let next_candidates_idx = sample_subsequent_candidates::<R, _>(
             &dists,
             F::cast(candidates_per_round),
-            rng.gen_range(0..std::u64::MAX),
+            rng.gen_range(0..u64::MAX),
         );
 
         // Append the newly generated candidates to the current cadidates, breaking out of the loop
@@ -191,6 +191,7 @@ fn k_means_para<R: Rng, F: Float, D: Distance<F>>(
 
 /// Generate candidate centroids by sampling each observation in parallel using a seedable RNG in
 /// every thread. Average number of generated candidates should equal `multiplier`.
+#[allow(clippy::extra_unused_type_parameters)]
 fn sample_subsequent_candidates<R: Rng, F: Float>(
     dists: &Array1<F>,
     multiplier: F,

algorithms/linfa-clustering/src/lib.rs

@@ -16,7 +16,7 @@
 //! * [K-Means](KMeans)
 //! * [DBSCAN](Dbscan)
 //! * [Approximated DBSCAN](AppxDbscan) (Currently an alias for DBSCAN, due to its superior
-//! performance)
+//!     performance)
 //! * [Gaussian-Mixture-Model](GaussianMixtureModel)
 //! * [OPTICS](OpticsAnalysis)
 //!

algorithms/linfa-clustering/src/optics/algorithm.rs

@@ -84,6 +84,7 @@ impl<F: Float> PartialEq for Sample<F> {
     }
 }
 
+#[allow(clippy::non_canonical_partial_ord_impl)]
 impl<F: Float> PartialOrd for Sample<F> {
     fn partial_cmp(&self, other: &Self) -> Option<Ordering> {
         self.reachability_distance

algorithms/linfa-elasticnet/examples/elasticnet_cv.rs

@@ -6,7 +6,7 @@ fn main() -> Result<()> {
     let mut dataset = linfa_datasets::diabetes();
 
     // parameters to compare
-    let ratios = vec![0.1, 0.2, 0.5, 0.7, 1.0];
+    let ratios = &[0.1, 0.2, 0.5, 0.7, 1.0];
 
     // create a model for each parameter
     let models = ratios

algorithms/linfa-elasticnet/src/lib.rs

@@ -35,8 +35,7 @@ pub use hyperparams::{
 ///
 /// See also:
 /// * [Talk on Fast Regularization Paths](https://web.stanford.edu/~hastie/TALKS/glmnet.pdf)
-/// * [Regularization Paths for Generalized Linear Models via Coordinate
-/// Descent](http://www.jstatsoft.org/v33/i01/paper)
+/// * [Regularization Paths for Generalized Linear Models via Coordinate Descent](http://www.jstatsoft.org/v33/i01/paper)
 #[derive(Debug, Clone)]
 pub struct ElasticNet<F> {
     hyperplane: Array1<F>,

algorithms/linfa-ftrl/examples/winequality_ftrl.rs

@@ -6,7 +6,7 @@ use rand::{rngs::SmallRng, SeedableRng};
 fn main() -> Result<()> {
     // Read the data
     let (train, valid) = linfa_datasets::winequality()
-        .map_targets(|v| if *v > 6 { true } else { false })
+        .map_targets(|v| *v > 6)
         .split_with_ratio(0.9);
 
     let params = Ftrl::params()

algorithms/linfa-ftrl/src/algorithm.rs

@@ -10,7 +10,7 @@ use rand::Rng;
 /// Simplified `Result` using [`FtrlError`](crate::FtrlError) as error type
 pub type Result<T> = std::result::Result<T, FtrlError>;
 
-impl<'a, F, R, D, T> FitWith<'a, ArrayBase<D, Ix2>, T, FtrlError> for FtrlValidParams<F, R>
+impl<F, R, D, T> FitWith<'_, ArrayBase<D, Ix2>, T, FtrlError> for FtrlValidParams<F, R>
 where
     F: Float,
     R: Rng + Clone,
@@ -255,7 +255,7 @@ mod test {
         let gradient: f64 = 0.5;
         let n: f64 = 0.11;
         let alpha = 0.5;
-        let expected_result = (((0.11 + 0.25) as f64).sqrt() - (0.11 as f64).sqrt()) / 0.5;
+        let expected_result = ((0.11f64 + 0.25).sqrt() - 0.11f64.sqrt()) / 0.5;
         let result = calculate_weight_in_average(n, gradient, alpha);
         assert_abs_diff_eq!(result, expected_result)
     }
@@ -302,7 +302,7 @@ mod test {
         let sigma = model.calculate_sigma(gradient.view());
         model.update_params(gradient.clone(), sigma.clone());
         let expected_z = initial_z + &gradient - sigma * weights;
-        let expected_n = initial_n + &gradient.mapv(|grad| (grad as f64).powf(2.));
+        let expected_n = initial_n + &gradient.mapv(|grad: f64| grad.powf(2.));
         assert_abs_diff_eq!(model.z(), &expected_z, epsilon = 1e-1);
         assert_abs_diff_eq!(model.n(), &expected_n, epsilon = 1e-1)
     }

algorithms/linfa-ftrl/src/lib.rs

@@ -49,7 +49,7 @@ impl<F: Float> Ftrl<F> {
     /// The description can be found [here](https://static.googleusercontent.com/media/research.google.com/en//pubs/archive/41159.pdf)
     ///
     /// It requires data preprocessing done in the separate step.
-
+    ///
     /// Create default hyperparameters. Random number generator will default to rand_xoshiro::Xoshiro256Plus
     pub fn params() -> FtrlParams<F, Xoshiro256Plus> {
         FtrlParams::default_with_rng(Xoshiro256Plus::seed_from_u64(42))

algorithms/linfa-hierarchical/examples/irisflower.rs

@@ -17,7 +17,7 @@ fn main() -> Result<(), Box<dyn Error>> {
         .transform(kernel)?;
 
     for (id, target) in kernel.targets().iter().zip(dataset.targets().into_iter()) {
-        let name = match *target as usize {
+        let name = match *target {
             0 => "setosa",
             1 => "versicolor",
             2 => "virginica",

algorithms/linfa-hierarchical/src/lib.rs

@@ -65,9 +65,9 @@ impl<F: Float> ParamGuard for HierarchicalCluster<F> {
 
     fn check_ref(&self) -> std::result::Result<&Self::Checked, Self::Error> {
         match self.0.stopping {
-            Criterion::NumClusters(x) if x == 0 => Err(
-                HierarchicalError::InvalidStoppingCondition(self.0.stopping.clone()),
-            ),
+            Criterion::NumClusters(0) => Err(HierarchicalError::InvalidStoppingCondition(
+                self.0.stopping.clone(),
+            )),
             Criterion::Distance(x) if x.is_negative() || x.is_nan() || x.is_infinite() => Err(
                 HierarchicalError::InvalidStoppingCondition(self.0.stopping.clone()),
             ),

algorithms/linfa-kernel/Cargo.toml

@@ -26,7 +26,7 @@ features = ["std", "derive"]
 [dependencies]
 ndarray = "0.15"
 num-traits = "0.2"
-sprs = { version="0.11", default-features = false }
+sprs = { version = "=0.11.1", default-features = false }
 
 linfa = { version = "0.7.0", path = "../.." }
 linfa-nn = { version = "0.7.0", path = "../linfa-nn" }

algorithms/linfa-kernel/src/inner.rs

@@ -102,7 +102,7 @@ impl<F: Float> Inner for CsMat<F> {
     }
 }
 
-impl<'a, F: Float> Inner for CsMatView<'a, F> {
+impl<F: Float> Inner for CsMatView<'_, F> {
     type Elem = F;
 
     fn dot(&self, rhs: &ArrayView2<F>) -> Array2<F> {