Add gaussian mixture model bench (#68)

The benchmark for the gaussian mixture model was missing

linfa-clustering/Cargo.toml

@@ -53,3 +53,7 @@ harness = false
 [[bench]]
 name = "appx_dbscan"
 harness = false
+
+[[bench]]
+name = "gaussian_mixture"
+harness = false

linfa-clustering/benches/gaussian_mixture.rs

@@ -0,0 +1,48 @@
+use criterion::{
+    black_box, criterion_group, criterion_main, AxisScale, Criterion, ParameterizedBenchmark,
+    PlotConfiguration,
+};
+use linfa::traits::Fit;
+use linfa::Dataset;
+use linfa_clustering::{generate_blobs, GaussianMixtureModel};
+use ndarray::Array2;
+use ndarray_rand::rand::SeedableRng;
+use ndarray_rand::rand_distr::Uniform;
+use ndarray_rand::RandomExt;
+use rand_isaac::Isaac64Rng;
+
+fn gaussian_mixture_bench(c: &mut Criterion) {
+    let mut rng = Isaac64Rng::seed_from_u64(40);
+    let cluster_sizes = vec![10, 100, 1000, 10000];
+
+    let benchmark = ParameterizedBenchmark::new(
+        "gaussian_mixture",
+        move |bencher, &cluster_size| {
+            let n_clusters = 4;
+            let n_features = 3;
+            let centroids =
+                Array2::random_using((n_clusters, n_features), Uniform::new(-30., 30.), &mut rng);
+            let dataset = Dataset::from(generate_blobs(cluster_size, &centroids, &mut rng));
+            bencher.iter(|| {
+                black_box(
+                    GaussianMixtureModel::params(n_clusters)
+                        .with_rng(rng.clone())
+                        .with_tolerance(1e-3)
+                        .with_max_n_iterations(1000)
+                        .fit(&dataset)
+                        .expect("GMM fitting fail"),
+                )
+            });
+        },
+        cluster_sizes,
+    )
+    .plot_config(PlotConfiguration::default().summary_scale(AxisScale::Logarithmic));
+    c.bench("gaussian_mixture", benchmark);
+}
+
+criterion_group! {
+  name = benches;
+  config = Criterion::default();
+  targets = gaussian_mixture_bench
+}
+criterion_main!(benches);