Upgrade to ndarray 0.16

.github/workflows/checking.yml

@@ -10,7 +10,7 @@ jobs:
       fail-fast: false
       matrix:
         toolchain:
-          - 1.82.0
+          - 1.87.0
           - stable
           - nightly
         os:

.github/workflows/testing.yml

@@ -10,7 +10,7 @@ jobs:
       fail-fast: false
       matrix:
         toolchain:
-          - 1.82.0
+          - 1.87.0
           - stable
         os:
           - ubuntu-latest
@@ -35,7 +35,7 @@ jobs:
       fail-fast: false
       matrix:
         toolchain:
-          - 1.82.0
+          - 1.87.0
           - stable
         os:
           - ubuntu-latest

Cargo.toml

@@ -38,15 +38,15 @@ serde = ["serde_crate", "ndarray/serde"]
 [dependencies]
 num-traits = "0.2"
 rand = { version = "0.8", features = ["small_rng"] }
-approx = "0.4"
+approx = "0.5"
 
-ndarray = { version = "0.15", features = ["approx"] }
-ndarray-linalg = { version = "0.16", optional = true }
-sprs = { version = "=0.11.1", default-features = false }
+ndarray = { version = "0.16", features = ["approx"] }
+ndarray-linalg = { version = "0.17", optional = true }
+sprs = { version = "0.11", default-features = false }
 
-thiserror = "1.0"
+thiserror = "2.0"
 
-criterion = { version = "0.4.0", optional = true }
+criterion = { version = "0.5", optional = true }
 
 [dependencies.serde_crate]
 package = "serde"
@@ -56,17 +56,17 @@ default-features = false
 features = ["std", "derive"]
 
 [dev-dependencies]
-ndarray-rand = "0.14"
+ndarray-rand = "0.15"
 linfa-datasets = { path = "datasets", features = [
     "winequality",
     "iris",
     "diabetes",
     "generate",
 ] }
-statrs = "0.16.0"
+statrs = "0.18"
 
 [target.'cfg(not(windows))'.dependencies]
-pprof = { version = "0.11.0", features = [
+pprof = { version = "0.14", features = [
     "flamegraph",
     "criterion",
 ], optional = true }

algorithms/linfa-bayes/Cargo.toml

@@ -21,12 +21,14 @@ default-features = false
 features = ["std", "derive"]
 
 [dependencies]
-ndarray = { version = "0.15" , features = ["approx"]}
-ndarray-stats = "0.5"
-thiserror = "1.0"
+ndarray = { version = "0.16", features = ["approx"] }
+ndarray-stats = "0.6"
+thiserror = "2.0"
 
 linfa = { version = "0.7.1", path = "../.." }
 
 [dev-dependencies]
-approx = "0.4"
-linfa-datasets = { version = "0.7.1", path = "../../datasets", features = ["winequality"] }
+approx = "0.5"
+linfa-datasets = { version = "0.7.1", path = "../../datasets", features = [
+    "winequality",
+] }

algorithms/linfa-bayes/src/base_nb.rs

@@ -70,7 +70,7 @@ where
             .mapv(|x| x.exp())
             .sum_axis(Axis(1))
             .mapv(|x| x.ln())
-            .into_shape((n_samples, 1))
+            .into_shape_with_order((n_samples, 1))
             .unwrap();
 
         (log_prob_mat - log_prob_x, classes)

algorithms/linfa-bayes/src/lib.rs

@@ -89,7 +89,7 @@ pub(crate) fn filter<F: Float, L: Label + Ord>(
     let index = y
         .into_iter()
         .enumerate()
-        .filter(|&(_, y)| (*ycondition == *y))
+        .filter(|&(_, y)| *ycondition == *y)
         .map(|(i, _)| i)
         .collect::<Vec<_>>();
 

algorithms/linfa-clustering/Cargo.toml

@@ -35,29 +35,29 @@ default-features = false
 features = ["std", "derive"]
 
 [dependencies]
-ndarray = { version = "0.15", features = ["rayon", "approx"] }
-linfa-linalg = { version = "0.1", default-features = false }
-ndarray-linalg = { version = "0.16", optional = true }
-ndarray-rand = "0.14"
-ndarray-stats = "0.5"
+ndarray = { version = "0.16", features = ["rayon", "approx"] }
+linfa-linalg = { version = "0.2", default-features = false }
+ndarray-linalg = { version = "0.17", optional = true }
+ndarray-rand = "0.15"
+ndarray-stats = "0.6"
 num-traits = "0.2"
 rand_xoshiro = "0.6"
 space = "0.12"
-thiserror = "1.0"
+thiserror = "2.0"
 #partitions = "0.2.4" This one will break in a future version of Rust and has no replacement
 linfa = { version = "0.7.1", path = "../.." }
 linfa-nn = { version = "0.7.2", path = "../linfa-nn" }
 noisy_float = "0.2.0"
 
 [dev-dependencies]
-ndarray-npy = { version = "0.8", default-features = false }
+ndarray-npy = { version = "0.9", default-features = false }
 linfa-datasets = { version = "0.7.1", path = "../../datasets", features = [
     "generate",
 ] }
-criterion = "0.4.0"
+criterion = "0.5"
 serde_json = "1"
-approx = "0.4"
-lax = "0.15.0"
+approx = "0.5"
+lax = "0.17.0"
 linfa = { version = "0.7.1", path = "../..", features = ["benchmarks"] }
 
 [[bench]]

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

@@ -396,7 +396,7 @@ impl<F: Float> GaussianMixtureModel<F> {
         let n_clusters = matrix_chol.shape()[0];
         let log_diags = &matrix_chol
             .to_owned()
-            .into_shape((n_clusters, n_features * n_features))
+            .into_shape_with_order((n_clusters, n_features * n_features))
             .unwrap()
             .slice(s![.., ..; n_features+1])
             .to_owned()

algorithms/linfa-elasticnet/Cargo.toml

@@ -29,13 +29,13 @@ default-features = false
 features = ["std", "derive"]
 
 [dependencies]
-ndarray = { version = "0.15", features = ["approx"] }
-linfa-linalg = { version = "0.1", default-features = false }
-ndarray-linalg = { version = "0.16", optional = true }
+ndarray = { version = "0.16", features = ["approx"] }
+linfa-linalg = { version = "0.2", default-features = false }
+ndarray-linalg = { version = "0.17", optional = true }
 
 num-traits = "0.2"
-approx = "0.4"
-thiserror = "1.0"
+approx = "0.5"
+thiserror = "2.0"
 
 linfa = { version = "0.7.1", path = "../.." }
 
@@ -44,5 +44,5 @@ linfa-datasets = { version = "0.7.1", path = "../../datasets", features = [
     "diabetes",
     "linnerud",
 ] }
-ndarray-rand = "0.14"
+ndarray-rand = "0.15"
 rand_xoshiro = "0.6"

algorithms/linfa-ensemble/Cargo.toml

@@ -27,8 +27,8 @@ default-features = false
 features = ["std", "derive"]
 
 [dependencies]
-ndarray = { version = "0.15", features = ["rayon", "approx"] }
-ndarray-rand = "0.14"
+ndarray = { version = "0.16", features = ["rayon", "approx"] }
+ndarray-rand = "0.15"
 rand = "0.8.5"
 
 linfa = { version = "0.7.1", path = "../.." }

algorithms/linfa-ftrl/Cargo.toml

@@ -24,19 +24,19 @@ version = "1.0"
 features = ["derive"]
 
 [dependencies]
-ndarray = { version = "0.15", features = ["serde"] }
-ndarray-rand = "0.14.0"
-argmin = { version = "0.9.0", default-features = false }
-argmin-math = { version = "0.3", features = ["ndarray_v0_15-nolinalg"] }
-thiserror = "1.0"
+ndarray = { version = "0.16", features = ["serde"] }
+ndarray-rand = "0.15.0"
+argmin = { version = "0.11.0", default-features = false }
+argmin-math = { version = "0.5", features = ["ndarray_v0_16-nolinalg"] }
+thiserror = "2.0"
 rand = "0.8.5"
 rand_xoshiro = "0.6.0"
 
 linfa = { version = "0.7.1", path = "../.." }
 
 [dev-dependencies]
-criterion = "0.4.0"
-approx = "0.4"
+criterion = "0.5"
+approx = "0.5"
 linfa-datasets = { version = "0.7.1", path = "../../datasets", features = [
     "winequality",
 ] }

algorithms/linfa-hierarchical/Cargo.toml

@@ -10,11 +10,17 @@ license = "MIT OR Apache-2.0"
 repository = "https://github.com/rust-ml/linfa"
 readme = "README.md"
 
-keywords = ["hierachical", "agglomerative", "clustering", "machine-learning", "linfa"]
+keywords = [
+    "hierachical",
+    "agglomerative",
+    "clustering",
+    "machine-learning",
+    "linfa",
+]
 categories = ["algorithms", "mathematics", "science"]
 
 [dependencies]
-ndarray = { version = "0.15" }
+ndarray = { version = "0.16" }
 kodama = "0.2"
 thiserror = "1.0.25"
 
@@ -23,5 +29,7 @@ linfa-kernel = { version = "0.7.1", path = "../linfa-kernel" }
 
 [dev-dependencies]
 rand = "0.8"
-ndarray-rand = "0.14"
-linfa-datasets = { version = "0.7.1", path = "../../datasets", features = ["iris"] }
+ndarray-rand = "0.15"
+linfa-datasets = { version = "0.7.1", path = "../../datasets", features = [
+    "iris",
+] }

algorithms/linfa-ica/Cargo.toml

@@ -25,21 +25,21 @@ default-features = false
 features = ["std", "derive"]
 
 [dependencies]
-ndarray = { version = "0.15" }
-linfa-linalg = { version = "0.1", default-features = false }
-ndarray-linalg = { version = "0.16", optional = true }
-ndarray-rand = "0.14"
-ndarray-stats = "0.5"
+ndarray = { version = "0.16" }
+linfa-linalg = { version = "0.2", default-features = false }
+ndarray-linalg = { version = "0.17", optional = true }
+ndarray-rand = "0.15"
+ndarray-stats = "0.6"
 num-traits = "0.2"
 rand_xoshiro = "0.6"
-thiserror = "1.0"
+thiserror = "2.0"
 
 linfa = { version = "0.7.1", path = "../.." }
 
 [dev-dependencies]
-ndarray-npy = { version = "0.8", default-features = false }
+ndarray-npy = { version = "0.9", default-features = false }
 paste = "1.0"
-criterion = "0.4.0"
+criterion = "0.5"
 linfa = { version = "0.7.1", path = "../..", features = ["benchmarks"] }
 
 [[bench]]

algorithms/linfa-kernel/Cargo.toml

@@ -24,9 +24,9 @@ default-features = false
 features = ["std", "derive"]
 
 [dependencies]
-ndarray = "0.15"
+ndarray = "0.16"
 num-traits = "0.2"
-sprs = { version = "=0.11.1", default-features = false }
+sprs = { version = "0.11", default-features = false }
 
 linfa = { version = "0.7.1", path = "../.." }
 linfa-nn = { version = "0.7.2", path = "../linfa-nn" }

algorithms/linfa-linear/Cargo.toml

@@ -29,23 +29,23 @@ default-features = false
 features = ["std", "derive"]
 
 [dependencies]
-ndarray = { version = "0.15", features = ["approx"] }
-linfa-linalg = { version = "0.1", default-features = false }
-ndarray-linalg = { version = "0.16", optional = true }
+ndarray = { version = "0.16", features = ["approx"] }
+linfa-linalg = { version = "0.2", default-features = false }
+ndarray-linalg = { version = "0.17", optional = true }
 num-traits = "0.2"
-argmin = { version = "0.9.0", default-features = false }
-argmin-math = { version = "0.3", features = ["ndarray_v0_15-nolinalg"] }
-thiserror = "1.0"
+argmin = { version = "0.11.0", default-features = false }
+argmin-math = { version = "0.5", features = ["ndarray_v0_16-nolinalg"] }
+thiserror = "2.0"
 
 linfa = { version = "0.7.1", path = "../.." }
 
 [dev-dependencies]
 linfa-datasets = { version = "0.7.1", path = "../../datasets", features = [
     "diabetes",
 ] }
-approx = "0.4"
-criterion = "0.4.0"
-statrs = "0.16.0"
+approx = "0.5"
+criterion = "0.5"
+statrs = "0.18"
 linfa = { version = "0.7.1", path = "../..", features = ["benchmarks"] }
 
 [[bench]]

algorithms/linfa-linear/src/isotonic.rs

@@ -265,8 +265,7 @@ impl<F: Float, D: Data<Elem = F>> PredictInplace<ArrayBase<D, Ix2>, Array1<F>>
                 y[i] = y_min;
             } else {
                 let res = regressor.into_iter().position(|x| x >= &val);
-                if res.is_some() {
-                    let j = res.unwrap();
+                if let Some(j) = res {
                     if val <= regressor[j] && j < n {
                         let x_scale = (val - regressor[j - 1]) / (regressor[j] - regressor[j - 1]);
                         y[i] = response[j - 1] + x_scale * (response[j] - response[j - 1]);

algorithms/linfa-logistic/Cargo.toml

@@ -23,17 +23,17 @@ optional = true
 version = "1.0"
 
 [dependencies]
-ndarray = { version = "0.15", features = ["approx"] }
-ndarray-stats = "0.5.0"
+ndarray = { version = "0.16", features = ["approx"] }
+ndarray-stats = "0.6.0"
 num-traits = "0.2"
-argmin = { version = "0.9.0", default-features = false }
-argmin-math = { version = "0.3", features = ["ndarray_v0_15-nolinalg"] }
-thiserror = "1.0"
+argmin = { version = "0.11.0", default-features = false }
+argmin-math = { version = "0.5", features = ["ndarray_v0_16-nolinalg"] }
+thiserror = "2.0"
 
 linfa = { version = "0.7.1", path = "../.." }
 
 [dev-dependencies]
-approx = "0.4"
+approx = "0.5"
 linfa-datasets = { version = "0.7.1", path = "../../datasets", features = [
     "winequality",
 ] }

algorithms/linfa-logistic/src/hyperparams.rs

@@ -13,17 +13,15 @@ use serde_crate::{Deserialize, Serialize};
 #[cfg_attr(
     feature = "serde",
     derive(Serialize, Deserialize),
-    serde(crate = "serde_crate"),
-    serde(bound(deserialize = "D: Deserialize<'de>"))
+    serde(crate = "serde_crate")
 )]
 pub struct LogisticRegressionParams<F: Float, D: Dimension>(LogisticRegressionValidParams<F, D>);
 
 #[derive(Debug, Clone, PartialEq)]
 #[cfg_attr(
     feature = "serde",
     derive(Serialize, Deserialize),
-    serde(crate = "serde_crate"),
-    serde(bound(deserialize = "D: Deserialize<'de>"))
+    serde(crate = "serde_crate")
 )]
 pub struct LogisticRegressionValidParams<F: Float, D: Dimension> {
     pub(crate) alpha: F,