Merge branch 'rust-ml:master' into master

.github/workflows/checking.yml

@@ -10,7 +10,7 @@ jobs:
       fail-fast: false
       matrix:
         toolchain:
-          - 1.65.0
+          - 1.67.0
           - stable
           - nightly
         os:
@@ -39,4 +39,4 @@ jobs:
         run: cargo check --workspace --all-targets
 
       - name: Run cargo check (with serde)
-        run: cargo check --workspace --all-targets --features "linfa-clustering/serde linfa-ica/serde linfa-kernel/serde linfa-reduction/serde linfa-svm/serde linfa-elasticnet/serde linfa-pls/serde linfa-trees/serde linfa-nn/serde linfa-linear/serde linfa-preprocessing/serde linfa-bayes/serde"
+        run: cargo check --workspace --all-targets --features "linfa-clustering/serde linfa-ica/serde linfa-kernel/serde linfa-reduction/serde linfa-svm/serde linfa-elasticnet/serde linfa-pls/serde linfa-trees/serde linfa-nn/serde linfa-linear/serde linfa-preprocessing/serde linfa-bayes/serde linfa-logistic/serde linfa-ftrl/serde"

.github/workflows/codequality.yml

@@ -10,7 +10,7 @@ jobs:
     strategy:
       matrix:
         toolchain:
-          - 1.65.0
+          - 1.67.0
           - stable
 
     steps:

.github/workflows/testing.yml

@@ -10,7 +10,7 @@ jobs:
       fail-fast: false
       matrix:
         toolchain:
-          - 1.65.0
+          - 1.67.0
           - stable
         os:
           - ubuntu-latest
@@ -35,7 +35,7 @@ jobs:
       fail-fast: false
       matrix:
         toolchain:
-          - 1.65.0
+          - 1.67.0
           - stable
         os:
           - ubuntu-latest

CHANGELOG.md

@@ -1,3 +1,18 @@
+Version 0.7.0 - 2023-10-15
+========================
+Changes
+-----------
+ * add `array_from_gz_csv` and `array_from_csv` in `linfa-datasets`
+ * make Serde support in `linfa-linear`, `linfa-logistic`, and `linfa-ftrl` optional
+ * bump `argmin` to 0.8.1
+ * add Serde support to `linfa-preprocessing` and `linfa-bayes`
+ * make licenses follow SPDX 2.1 license expression standard
+
+Removals
+-----------
+ * Removed Approximate DBSCAN from `linfa-clustering` due to performance issues. It's now an alias to regular DBSCAN.
+ * Removed `partitions` dependency, which breaks in current versions of Rust.
+
 Version 0.6.1 - 2022-12-03
 ========================
 New Algorithms

Cargo.toml

@@ -1,6 +1,6 @@
 [package]
 name = "linfa"
-version = "0.6.1"
+version = "0.7.0"
 authors = [
     "Luca Palmieri <[email protected]>",
     "Lorenz Schmidt <[email protected]>",
@@ -9,7 +9,7 @@ authors = [
 ]
 description = "A Machine Learning framework for Rust"
 edition = "2018"
-license = "MIT/Apache-2.0"
+license = "MIT OR Apache-2.0"
 
 repository = "https://github.com/rust-ml/linfa"
 readme = "README.md"

algorithms/linfa-bayes/Cargo.toml

@@ -1,10 +1,10 @@
 [package]
 name = "linfa-bayes"
-version = "0.6.1"
+version = "0.7.0"
 authors = ["VasanthakumarV <[email protected]>"]
 description = "Collection of Naive Bayes Algorithms"
 edition = "2018"
-license = "MIT/Apache-2.0"
+license = "MIT OR Apache-2.0"
 repository = "https://github.com/rust-ml/linfa"
 readme = "README.md"
 keywords = ["factorization", "machine-learning", "linfa", "unsupervised"]
@@ -25,8 +25,8 @@ ndarray = { version = "0.15" , features = ["approx"]}
 ndarray-stats = "0.5"
 thiserror = "1.0"
 
-linfa = { version = "0.6.1", path = "../.." }
+linfa = { version = "0.7.0", path = "../.." }
 
 [dev-dependencies]
 approx = "0.4"
-linfa-datasets = { version = "0.6.1", path = "../../datasets", features = ["winequality"] }
+linfa-datasets = { version = "0.7.0", path = "../../datasets", features = ["winequality"] }

algorithms/linfa-clustering/Cargo.toml

@@ -1,14 +1,14 @@
 [package]
 name = "linfa-clustering"
-version = "0.6.1"
+version = "0.7.0"
 edition = "2018"
 authors = [
     "Luca Palmieri <[email protected]>",
     "xd009642 <[email protected]>",
     "Rémi Lafage <[email protected]>"
 ]
 description = "A collection of clustering algorithms"
-license = "MIT/Apache-2.0"
+license = "MIT OR Apache-2.0"
 
 repository = "https://github.com/rust-ml/linfa/"
 readme = "README.md"
@@ -38,18 +38,18 @@ rand_xoshiro = "0.6"
 space = "0.12"
 thiserror = "1.0"
 #partitions = "0.2.4" This one will break in a future version of Rust and has no replacement
-linfa = { version = "0.6.1", path = "../.." }
-linfa-nn = { version = "0.6.1", path = "../linfa-nn" }
+linfa = { version = "0.7.0", path = "../.." }
+linfa-nn = { version = "0.7.0", path = "../linfa-nn" }
 noisy_float = "0.2.0"
 
 [dev-dependencies]
 ndarray-npy = { version = "0.8", default-features = false }
-linfa-datasets = { version = "0.6.1", path = "../../datasets", features = ["generate"] }
+linfa-datasets = { version = "0.7.0", path = "../../datasets", features = ["generate"] }
 criterion = "0.4.0"
 serde_json = "1"
 approx = "0.4"
 lax = "0.15.0"
-linfa = { version = "0.6.0", path = "../..", features = ["benchmarks"] }
+linfa = { version = "0.7.0", path = "../..", features = ["benchmarks"] }
 
 [[bench]]
 name = "k_means"

algorithms/linfa-elasticnet/Cargo.toml

@@ -1,14 +1,14 @@
 [package]
 name = "linfa-elasticnet"
-version = "0.6.1"
+version = "0.7.0"
 authors = [
     "Paul Körbitz / Google <[email protected]>",
     "Lorenz Schmidt <[email protected]>"
 ]
 
 description = "A Machine Learning framework for Rust"
 edition = "2018"
-license = "MIT/Apache-2.0"
+license = "MIT OR Apache-2.0"
 
 repository = "https://github.com/rust-ml/linfa"
 readme = "README.md"
@@ -37,9 +37,9 @@ num-traits = "0.2"
 approx = "0.4"
 thiserror = "1.0"
 
-linfa = { version = "0.6.1", path = "../.." }
+linfa = { version = "0.7.0", path = "../.." }
 
 [dev-dependencies]
-linfa-datasets = { version = "0.6.1", path = "../../datasets", features = ["diabetes", "linnerud"] }
+linfa-datasets = { version = "0.7.0", path = "../../datasets", features = ["diabetes", "linnerud"] }
 ndarray-rand = "0.14"
 rand_xoshiro = "0.6"

algorithms/linfa-ftrl/Cargo.toml

@@ -1,34 +1,43 @@
 [package]
 name = "linfa-ftrl"
-version = "0.6.1"
+version = "0.7.0"
 authors = ["Liudmyla Kyrashchuk <[email protected]>"]
 
 description = "A Machine Learning framework for Rust"
 edition = "2018"
-license = "MIT/Apache-2.0"
+license = "MIT OR Apache-2.0"
 
 repository = "https://github.com/rust-ml/linfa"
 readme = "README.md"
 
 keywords = ["machine-learning", "linfa", "ai", "ml", "ftrl"]
 categories = ["algorithms", "mathematics", "science"]
 
+[features]
+serde = ["serde_crate", "linfa/serde", "ndarray/serde", "argmin/serde1"]
+
+[dependencies.serde_crate]
+package = "serde"
+optional = true
+version = "1.0"
+features = ["derive"]
+
 [dependencies]
 ndarray = { version = "0.15.4", features = ["serde"]}
 ndarray-rand = "0.14.0"
-argmin = { version = "0.4.7", features = ["ndarray", "ndarray-rand"]}
+argmin = { version = "0.8.1", default-features = false }
+argmin-math = { version = "0.3", features = ["ndarray_v0_15-nolinalg"] }
 thiserror = "1.0"
 rand = "0.8.5"
 rand_xoshiro = "0.6.0"
-serde = {version = "1.0.137", features = ["derive"]}
 
-linfa = { version = "0.6.1", path = "../.."}
+linfa = { version = "0.7.0", path = "../.."}
 
 [dev-dependencies]
 criterion = "0.4.0"
 approx = "0.4"
-linfa-datasets = { version = "0.6.1", path = "../../datasets", features = ["winequality"] }
-linfa = { version = "0.6.1", path = "../..", features = ["benchmarks"] }
+linfa-datasets = { version = "0.7.0", path = "../../datasets", features = ["winequality"] }
+linfa = { version = "0.7.0", path = "../..", features = ["benchmarks"] }
 
 [[bench]]
 name = "ftrl"

algorithms/linfa-ftrl/src/hyperparams.rs

@@ -1,15 +1,26 @@
 use crate::error::FtrlError;
 use linfa::{Float, ParamGuard};
 use rand::Rng;
-use serde::{Deserialize, Serialize};
-
-#[derive(Clone, Debug, Serialize, Deserialize, PartialEq)]
+#[cfg(feature = "serde")]
+use serde_crate::{Deserialize, Serialize};
+
+#[derive(Debug, Clone, PartialEq)]
+#[cfg_attr(
+    feature = "serde",
+    derive(Serialize, Deserialize),
+    serde(crate = "serde_crate")
+)]
 pub struct FtrlParams<F: Float, R: Rng>(pub(crate) FtrlValidParams<F, R>);
 
 /// A verified hyper-parameter set ready for the estimation of a Follow the regularized leader - proximal model
 ///
 /// See [`FtrlParams`](crate::FtrlParams) for more information.
-#[derive(Clone, Debug, Serialize, Deserialize, PartialEq)]
+#[derive(Debug, Clone, PartialEq)]
+#[cfg_attr(
+    feature = "serde",
+    derive(Serialize, Deserialize),
+    serde(crate = "serde_crate")
+)]
 pub struct FtrlValidParams<F: Float, R: Rng> {
     pub(crate) alpha: F,
     pub(crate) beta: F,

algorithms/linfa-ftrl/src/lib.rs

@@ -13,9 +13,15 @@ use ndarray::Array1;
 use ndarray_rand::RandomExt;
 use rand::{distributions::Uniform, Rng};
 use rand_xoshiro::{rand_core::SeedableRng, Xoshiro256Plus};
-use serde::{Deserialize, Serialize};
+#[cfg(feature = "serde")]
+use serde_crate::{Deserialize, Serialize};
 
-#[derive(Clone, Debug, Serialize, Deserialize)]
+#[derive(Debug, Clone)]
+#[cfg_attr(
+    feature = "serde",
+    derive(Serialize, Deserialize),
+    serde(crate = "serde_crate")
+)]
 pub struct Ftrl<F: Float> {
     /// FTRL (Follow The Regularized Leader - proximal) is a linear model for CTR prediction in online learning settings.
     /// It stores z and n values, which are later used to calculate weights at incremental model fit and during prediction.

algorithms/linfa-hierarchical/Cargo.toml

@@ -1,11 +1,11 @@
 [package]
 name = "linfa-hierarchical"
-version = "0.6.1"
+version = "0.7.0"
 authors = ["Lorenz Schmidt <[email protected]>"]
 edition = "2018"
 
 description = "Agglomerative Hierarchical clustering"
-license = "MIT/Apache-2.0"
+license = "MIT OR Apache-2.0"
 
 repository = "https://github.com/rust-ml/linfa"
 readme = "README.md"
@@ -16,12 +16,12 @@ categories = ["algorithms", "mathematics", "science"]
 [dependencies]
 ndarray = { version = "0.15" }
 kodama = "0.2"
-thiserror = "=1.0.25"
+thiserror = "1.0.25"
 
-linfa = { version = "0.6.1", path = "../.." }
-linfa-kernel = { version = "0.6.1", path = "../linfa-kernel" }
+linfa = { version = "0.7.0", path = "../.." }
+linfa-kernel = { version = "0.7.0", path = "../linfa-kernel" }
 
 [dev-dependencies]
 rand = "0.8"
 ndarray-rand = "0.14"
-linfa-datasets = { version = "0.6.1", path = "../../datasets", features = ["iris"] }
+linfa-datasets = { version = "0.7.0", path = "../../datasets", features = ["iris"] }

algorithms/linfa-ica/Cargo.toml

@@ -1,10 +1,10 @@
 [package]
 name = "linfa-ica"
-version = "0.6.1"
+version = "0.7.0"
 authors = ["VasanthakumarV <[email protected]>"]
 description = "A collection of Independent Component Analysis (ICA) algorithms"
 edition = "2018"
-license = "MIT/Apache-2.0"
+license = "MIT OR Apache-2.0"
 
 repository = "https://github.com/rust-ml/linfa"
 readme = "README.md"
@@ -34,13 +34,13 @@ num-traits = "0.2"
 rand_xoshiro = "0.6"
 thiserror = "1.0"
 
-linfa = { version = "0.6.1", path = "../.." }
+linfa = { version = "0.7.0", path = "../.." }
 
 [dev-dependencies]
 ndarray-npy = { version = "0.8", default-features = false }
 paste = "1.0"
 criterion = "0.4.0"
-linfa = { version = "0.6.0", path = "../..", features = ["benchmarks"] }
+linfa = { version = "0.7.0", path = "../..", features = ["benchmarks"] }
 
 [[bench]]
 name = "fast_ica"

algorithms/linfa-kernel/Cargo.toml

@@ -1,10 +1,10 @@
 [package]
 name = "linfa-kernel"
-version = "0.6.1"
+version = "0.7.0"
 authors = ["Lorenz Schmidt <[email protected]>"]
 description = "Kernel methods for non-linear algorithms"
 edition = "2018"
-license = "MIT/Apache-2.0"
+license = "MIT OR Apache-2.0"
 
 repository = "https://github.com/rust-ml/linfa"
 readme = "README.md"
@@ -28,5 +28,5 @@ ndarray = "0.15"
 num-traits = "0.2"
 sprs = { version="0.11", default-features = false }
 
-linfa = { version = "0.6.1", path = "../.." }
-linfa-nn = { version = "0.6.1", path = "../linfa-nn" }
+linfa = { version = "0.7.0", path = "../.." }
+linfa-nn = { version = "0.7.0", path = "../linfa-nn" }

algorithms/linfa-linear/Cargo.toml

@@ -1,14 +1,14 @@
 [package]
 name = "linfa-linear"
-version = "0.6.1"
+version = "0.7.0"
 authors = [
     "Paul Körbitz / Google <[email protected]>",
     "VasanthakumarV <[email protected]>"
 ]
 
 description = "A Machine Learning framework for Rust"
 edition = "2018"
-license = "MIT/Apache-2.0"
+license = "MIT OR Apache-2.0"
 
 repository = "https://github.com/rust-ml/linfa"
 readme = "README.md"
@@ -32,18 +32,18 @@ ndarray = { version = "0.15", features = ["approx"] }
 linfa-linalg = { version = "0.1", default-features = false }
 ndarray-linalg = { version = "0.15", optional = true }
 num-traits = "0.2"
-argmin = { version = "0.7", default-features = false }
-argmin-math = { version = "0.2", features = ["ndarray_v0_15-nolinalg"] }
+argmin = { version = "0.8.1", default-features = false }
+argmin-math = { version = "0.3", features = ["ndarray_v0_15-nolinalg"] }
 thiserror = "1.0"
 
-linfa = { version = "0.6.1", path = "../.." }
+linfa = { version = "0.7.0", path = "../.." }
 
 [dev-dependencies]
-linfa-datasets = { version = "0.6.1", path = "../../datasets", features = ["diabetes"] }
+linfa-datasets = { version = "0.7.0", path = "../../datasets", features = ["diabetes"] }
 approx = "0.4"
 criterion = "0.4.0" 
 statrs = "0.16.0"
-linfa = { version = "0.6.0", path = "../..", features = ["benchmarks"] }
+linfa = { version = "0.7.0", path = "../..", features = ["benchmarks"] }
 
 [[bench]]
 name = "ols_bench"