kmeans 2.0.2

Small and fast library for k-means clustering calculations.
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326

use crate::api::DistanceFunction;
use crate::memory::*;
use crate::{KMeans, KMeansConfig, KMeansState};
use rand::prelude::*;
use rayon::prelude::*;
use std::ops::{DerefMut, Range};
use std::simd::Simd;

struct BatchInfo {
    start_idx: usize,
    batch_size: usize,
}
impl BatchInfo {
    fn gen_range(&self, stride: usize) -> Range<usize> {
        Range {
            start: (self.start_idx * stride),
            end: (self.start_idx * stride + self.batch_size * stride),
        }
    }
}

pub(crate) struct Minibatch<T, const LANES: usize, D>
where
    T: Primitive,
    D: DistanceFunction<T, LANES>,
{
    _p: std::marker::PhantomData<(T, D)>,
}

impl<T, const LANES: usize, D> Minibatch<T, LANES, D>
where
    T: Primitive,
    Simd<T, LANES>: SupportedSimdArray<T, LANES>,
    D: DistanceFunction<T, LANES>,
{
    fn update_cluster_assignments(
        data: &KMeans<T, LANES, D>, state: &mut KMeansState<T>, batch: &BatchInfo, shuffled_samples: &[T], limit_k: Option<usize>,
    ) {
        let centroids = &state.centroids;
        let k = limit_k.unwrap_or(state.k);

        // TODO: Switch to par_chunks_exact, when that is merged in rayon (https://github.com/rayon-rs/rayon/pull/629).
        // par_chunks() works, because sample-dimensions are manually padded, so that there is no remainder

        // manually calculate work-packet size, because rayon does not do static scheduling (which is more apropriate here)
        let work_packet_size = batch.batch_size / rayon::current_num_threads();
        shuffled_samples[batch.gen_range(data.p_samples.stride)]
            .par_chunks_exact(data.p_samples.stride)
            .with_min_len(work_packet_size)
            .zip(state.assignments[batch.gen_range(1)].par_iter_mut())
            .zip(state.centroid_distances[batch.gen_range(1)].par_iter_mut())
            .for_each(|((s, assignment), centroid_dist)| {
                let (best_idx, best_dist) = centroids
                    .chunks_exact_stride()
                    .take(k)
                    .map(|c| data.distance_fn.distance(s, c))
                    .enumerate()
                    .min_by(|(_, d0), (_, d1)| d0.partial_cmp(d1).unwrap())
                    .unwrap();
                *assignment = best_idx;
                *centroid_dist = best_dist;
            });
    }

    fn update_centroids(data: &KMeans<T, LANES, D>, state: &mut KMeansState<T>, batch: &BatchInfo, shuffled_samples: &[T]) {
        let centroid_frequency = &mut state.centroid_frequency;
        let centroids = &mut state.centroids;
        let assignments = &state.assignments;

        shuffled_samples[batch.gen_range(data.p_samples.stride)]
            .chunks_exact(data.p_samples.stride)
            .zip(assignments[batch.gen_range(1)].iter().cloned())
            .for_each(|(sample, assignment)| {
                centroid_frequency[assignment] += 1;
                let learn_rate = T::one() / T::from(centroid_frequency[assignment]).unwrap();
                let inv_learn_rate = T::one() - learn_rate;
                centroids
                    .bfr
                    .iter_mut()
                    .skip(assignment * data.p_samples.stride)
                    .take(data.p_samples.stride)
                    .zip(sample.iter().cloned())
                    .for_each(|(c, s)| {
                        *c = inv_learn_rate * *c + learn_rate * s;
                    });
            });
    }

    fn shuffle_samples(data: &KMeans<T, LANES, D>, config: &KMeansConfig<'_, T>) -> (Vec<usize>, StrideBuffer<T>) {
        let mut idxs: Vec<usize> = (0..data.sample_cnt).collect();
        idxs.shuffle(config.rnd.borrow_mut().deref_mut());

        let mut shuffled_samples = StrideBuffer::new::<LANES>(data.sample_cnt, data.sample_dims);
        shuffled_samples.iter_mut().zip(idxs.iter()).for_each(|(dst, src_idx)| {
            dst.copy_from_slice(&data.p_samples[*src_idx]);
        });

        (idxs, shuffled_samples)
    }

    fn unshuffle_state(shuffle_idxs: &[usize], state: &mut KMeansState<T>) {
        for (from, to) in shuffle_idxs.iter().cloned().enumerate() {
            state.assignments.swap(from, to);
            state.centroid_distances.swap(from, to);
        }
    }

    #[inline(always)]
    pub fn calculate<F>(
        data: &KMeans<T, LANES, D>, batch_size: usize, k: usize, max_iter: usize, init: F, config: &KMeansConfig<'_, T>,
    ) -> KMeansState<T>
    where
        for<'c> F: FnOnce(&KMeans<T, LANES, D>, &mut KMeansState<T>, &KMeansConfig<'c, T>),
    {
        assert!(k <= data.sample_cnt);
        assert!(batch_size <= data.sample_cnt);

        // Copy and shuffle sample_data, then only take consecutive blocks (with batch_size) from there
        let (shuffle_idxs, shuffled_samples) = Self::shuffle_samples(data, config);

        let mut state = KMeansState::new::<LANES>(data.sample_cnt, data.sample_dims, k);
        state.distsum = T::infinity();

        // Initialize clusters and notify subscriber
        init(data, &mut state, config);
        (config.init_done)(&state);
        let mut abort_strategy = config.abort_strategy.create_logic();

        // Update cluster assignments for all samples, to get rid of the INFINITES in centroid_distances
        Self::update_cluster_assignments(
            data,
            &mut state,
            &BatchInfo {
                start_idx: 0,
                batch_size: data.sample_cnt,
            },
            &shuffled_samples.bfr, //TODO
            None,
        );

        for i in 1..=max_iter {
            // Only shuffle a beginning index for a consecutive block within the shuffled samples as batch
            let batch = BatchInfo {
                batch_size,
                start_idx: config.rnd.borrow_mut().gen_range(0..data.sample_cnt - batch_size),
            };

            Self::update_cluster_assignments(data, &mut state, &batch, &shuffled_samples.bfr, None);
            let new_distsum = state.centroid_distances.iter().cloned().sum();
            Self::update_centroids(data, &mut state, &batch, &shuffled_samples.bfr);

            // Notify subscriber about finished iteration
            (config.iteration_done)(&state, i, new_distsum);
            if !abort_strategy.next(new_distsum) {
                break;
            }
            state.distsum = new_distsum;
        }

        // Unshuffle state, so that all sample - x mappings are aligned to the original data.p_samples input
        Self::unshuffle_state(&shuffle_idxs, &mut state);
        data.update_cluster_assignments(&mut state, None);

        let (assignments, centroid_frequency, centroid_distances, distsum) = (
            &state.assignments, &mut state.centroid_frequency, &mut state.centroid_distances, &mut state.distsum,
        );
        let mut non_empty_clusters = state.k;
        rayon::scope(|s| {
            s.spawn(|_| {
                non_empty_clusters -= data.update_cluster_frequencies(assignments, centroid_frequency);
            });
            s.spawn(|_| {
                *distsum = centroid_distances.iter().cloned().sum();
            });
        });
        state
    }
}

#[cfg(test)]
mod tests {
    use super::*;
    use crate::helpers::testing::{assert_kmeans_result_eq, KMeansShouldResult};
    use crate::{AbortStrategy, EuclideanDistance};

    #[test]
    fn iris_dataset_f64() {
        let samples = vec![
            1.4f64, 0.2, 1.4, 0.2, 1.3, 0.2, 1.5, 0.2, 1.4, 0.2, 1.7, 0.4, 1.4, 0.3, 1.5, 0.2, 1.4, 0.2, 1.5, 0.1, 1.5, 0.2, 1.6, 0.2, 1.4,
            0.1, 1.1, 0.1, 1.2, 0.2, 1.5, 0.4, 1.3, 0.4, 1.4, 0.3, 1.7, 0.3, 1.5, 0.3, 1.7, 0.2, 1.5, 0.4, 1.0, 0.2, 1.7, 0.5, 1.9, 0.2,
            1.6, 0.2, 1.6, 0.4, 1.5, 0.2, 1.4, 0.2, 1.6, 0.2, 1.6, 0.2, 1.5, 0.4, 1.5, 0.1, 1.4, 0.2, 1.5, 0.2, 1.2, 0.2, 1.3, 0.2, 1.4,
            0.1, 1.3, 0.2, 1.5, 0.2, 1.3, 0.3, 1.3, 0.3, 1.3, 0.2, 1.6, 0.6, 1.9, 0.4, 1.4, 0.3, 1.6, 0.2, 1.4, 0.2, 1.5, 0.2, 1.4, 0.2,
            4.7, 1.4, 4.5, 1.5, 4.9, 1.5, 4.0, 1.3, 4.6, 1.5, 4.5, 1.3, 4.7, 1.6, 3.3, 1.0, 4.6, 1.3, 3.9, 1.4, 3.5, 1.0, 4.2, 1.5, 4.0,
            1.0, 4.7, 1.4, 3.6, 1.3, 4.4, 1.4, 4.5, 1.5, 4.1, 1.0, 4.5, 1.5, 3.9, 1.1, 4.8, 1.8, 4.0, 1.3, 4.9, 1.5, 4.7, 1.2, 4.3, 1.3,
            4.4, 1.4, 4.8, 1.4, 5.0, 1.7, 4.5, 1.5, 3.5, 1.0, 3.8, 1.1, 3.7, 1.0, 3.9, 1.2, 5.1, 1.6, 4.5, 1.5, 4.5, 1.6, 4.7, 1.5, 4.4,
            1.3, 4.1, 1.3, 4.0, 1.3, 4.4, 1.2, 4.6, 1.4, 4.0, 1.2, 3.3, 1.0, 4.2, 1.3, 4.2, 1.2, 4.2, 1.3, 4.3, 1.3, 3.0, 1.1, 4.1, 1.3,
            6.0, 2.5, 5.1, 1.9, 5.9, 2.1, 5.6, 1.8, 5.8, 2.2, 6.6, 2.1, 4.5, 1.7, 6.3, 1.8, 5.8, 1.8, 6.1, 2.5, 5.1, 2.0, 5.3, 1.9, 5.5,
            2.1, 5.0, 2.0, 5.1, 2.4, 5.3, 2.3, 5.5, 1.8, 6.7, 2.2, 6.9, 2.3, 5.0, 1.5, 5.7, 2.3, 4.9, 2.0, 6.7, 2.0, 4.9, 1.8, 5.7, 2.1,
            6.0, 1.8, 4.8, 1.8, 4.9, 1.8, 5.6, 2.1, 5.8, 1.6, 6.1, 1.9, 6.4, 2.0, 5.6, 2.2, 5.1, 1.5, 5.6, 1.4, 6.1, 2.3, 5.6, 2.4, 5.5,
            1.8, 4.8, 1.8, 5.4, 2.1, 5.6, 2.4, 5.1, 2.3, 5.1, 1.9, 5.9, 2.3, 5.7, 2.5, 5.2, 2.3, 5.0, 1.9, 5.2, 2.0, 5.4, 2.3, 5.1, 1.8,
        ];

        let kmean: KMeans<f64, 8, _> = KMeans::new(&samples, 150, 2, EuclideanDistance);
        let rnd = rand::rngs::StdRng::seed_from_u64(3);
        let conf = KMeansConfig::build()
            .random_generator(rnd)
            .abort_strategy(AbortStrategy::NoImprovementForXIterations {
                x: 5,
                threshold: 0.0005f64,
                abort_on_negative: true,
            })
            .build();
        let res = kmean.kmeans_minibatch(30, 3, 100, KMeans::init_kmeanplusplus, &conf);

        // SHOULD solution
        let should = KMeansShouldResult {
            distsum: 31.691483430123924,
            sample_dims: 2,
            assignments: vec![
                1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
                1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
                0, 2, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 2, 2, 2, 2, 2, 2, 0, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 0, 2, 2, 2,
                0, 2, 2, 0, 0, 2, 2, 2, 2, 2, 0, 2, 2, 2, 2, 0, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
            ],
            centroid_distances: vec![
                0.008367214532872126, 0.008367214532872126, 0.03307309688581347, 0.0036613321799307407, 0.008367214532872126,
                0.07248486159169505, 0.00748486159169566, 0.0036613321799307407, 0.008367214532872126, 0.02454368512110721,
                0.0036613321799307407, 0.018955449826989393, 0.029249567474048596, 0.16336721453287262, 0.07777897923875488,
                0.021896626297577815, 0.051308391003460546, 0.00748486159169566, 0.05336721453287151, 0.0027789792387542737,
                0.05424956747404797, 0.021896626297577815, 0.2271907439446376, 0.11160250865051857, 0.18483780276816517,
                0.018955449826989393, 0.037190743944636465, 0.0036613321799307407, 0.008367214532872126, 0.018955449826989393,
                0.018955449826989393, 0.021896626297577815, 0.02454368512110721, 0.008367214532872126, 0.0036613321799307407,
                0.07777897923875488, 0.03307309688581347, 0.029249567474048596, 0.03307309688581347, 0.0036613321799307407,
                0.032190743944637, 0.032190743944637, 0.03307309688581347, 0.1354260380622835, 0.20307309688581224, 0.00748486159169566,
                0.018955449826989393, 0.008367214532872126, 0.0036613321799307407, 0.008367214532872126, 0.11522168220591258,
                0.02902449910732018, 0.30057379488196956, 0.1404329498115445, 0.06691182305098226, 0.029869569529855745,
                0.15437661178337703, 1.2864892878397125, 0.06775689347351782, 0.21212309065661458, 0.9022639357270366,
                0.035362527276333384, 0.2917005554453479, 0.11522168220591258, 0.5888836540368955, 0.0015597103749257503,
                0.02902449910732018, 0.22958787938901032, 0.02902449910732018, 0.30339069629041787, 0.3514188653045034, 0.1404329498115445,
                0.30057379488196956, 0.15606675262844816, 0.014094921642531275, 0.0015597103749257503, 0.19310900614957455,
                0.4976160484030958, 0.02902449910732018, 0.9022639357270366, 0.40550337234675576, 0.5980385836143609, 0.25296816107915016,
                0.5381776147959252, 0.02902449910732018, 0.05860196389605243, 0.1247991469946448, 0.011982245586193527,
                0.07832027375520695, 0.1404329498115445, 0.04240478079746134, 0.05733435826225004, 0.17085548502281234, 1.2864892878397125,
                0.03620759769886895, 0.06663013291013677, 0.03620759769886895, 0.014094921642531275, 1.9424047807974576,
                0.07832027375520695, 0.2971061862244855, 0.34371332908163893, 0.058356186224487565, 0.0787133290816339,
                0.03549904336734527, 0.88460618622448, 0.10817942868478458, 0.48496332908162676, 0.09478475765306044, 0.37514190051019847,
                0.3188919005102101, 0.1597847576530653, 0.026213329081634155, 0.4408561862244964, 0.4196061862244948, 0.18049904336734995,
                0.10067761479592055, 1.0978204719387654, 1.5890704719387636, 0.4184611188256314, 0.05264190051020309, 0.5828204719387825,
                1.0874633290816231, 0.44930618924816623, 0.002284757653060791, 0.1908561862244871, 0.3514188653045034, 0.44930618924816623,
                0.0042490433673475075, 0.24442761479591799, 0.22407047193877133, 0.5533561862244832, 0.01942761479591874,
                0.5563484427692932, 0.45799904336734926, 0.244784757653056, 0.10978475765306102, 0.10067761479592055, 0.3514188653045034,
                0.06817761479592066, 0.10978475765306102, 0.3644276147959236, 0.34371332908163893, 0.10871332908162987,
                0.18299904336734557, 0.2624633290816364, 0.46567761479592523, 0.21692761479592287, 0.11853475765306296,
                0.38853475765306766,
            ],
            centroids: vec![
                4.360563380281689, 1.402112676056339, 1.4735294117647069, 0.25441176470588234, 5.659821428571433, 2.074107142857144,
            ],
        };

        assert_kmeans_result_eq(should, res);
    }

    #[test]
    fn iris_dataset_f32() {
        let samples = vec![
            1.4f32, 0.2, 1.4, 0.2, 1.3, 0.2, 1.5, 0.2, 1.4, 0.2, 1.7, 0.4, 1.4, 0.3, 1.5, 0.2, 1.4, 0.2, 1.5, 0.1, 1.5, 0.2, 1.6, 0.2, 1.4,
            0.1, 1.1, 0.1, 1.2, 0.2, 1.5, 0.4, 1.3, 0.4, 1.4, 0.3, 1.7, 0.3, 1.5, 0.3, 1.7, 0.2, 1.5, 0.4, 1.0, 0.2, 1.7, 0.5, 1.9, 0.2,
            1.6, 0.2, 1.6, 0.4, 1.5, 0.2, 1.4, 0.2, 1.6, 0.2, 1.6, 0.2, 1.5, 0.4, 1.5, 0.1, 1.4, 0.2, 1.5, 0.2, 1.2, 0.2, 1.3, 0.2, 1.4,
            0.1, 1.3, 0.2, 1.5, 0.2, 1.3, 0.3, 1.3, 0.3, 1.3, 0.2, 1.6, 0.6, 1.9, 0.4, 1.4, 0.3, 1.6, 0.2, 1.4, 0.2, 1.5, 0.2, 1.4, 0.2,
            4.7, 1.4, 4.5, 1.5, 4.9, 1.5, 4.0, 1.3, 4.6, 1.5, 4.5, 1.3, 4.7, 1.6, 3.3, 1.0, 4.6, 1.3, 3.9, 1.4, 3.5, 1.0, 4.2, 1.5, 4.0,
            1.0, 4.7, 1.4, 3.6, 1.3, 4.4, 1.4, 4.5, 1.5, 4.1, 1.0, 4.5, 1.5, 3.9, 1.1, 4.8, 1.8, 4.0, 1.3, 4.9, 1.5, 4.7, 1.2, 4.3, 1.3,
            4.4, 1.4, 4.8, 1.4, 5.0, 1.7, 4.5, 1.5, 3.5, 1.0, 3.8, 1.1, 3.7, 1.0, 3.9, 1.2, 5.1, 1.6, 4.5, 1.5, 4.5, 1.6, 4.7, 1.5, 4.4,
            1.3, 4.1, 1.3, 4.0, 1.3, 4.4, 1.2, 4.6, 1.4, 4.0, 1.2, 3.3, 1.0, 4.2, 1.3, 4.2, 1.2, 4.2, 1.3, 4.3, 1.3, 3.0, 1.1, 4.1, 1.3,
            6.0, 2.5, 5.1, 1.9, 5.9, 2.1, 5.6, 1.8, 5.8, 2.2, 6.6, 2.1, 4.5, 1.7, 6.3, 1.8, 5.8, 1.8, 6.1, 2.5, 5.1, 2.0, 5.3, 1.9, 5.5,
            2.1, 5.0, 2.0, 5.1, 2.4, 5.3, 2.3, 5.5, 1.8, 6.7, 2.2, 6.9, 2.3, 5.0, 1.5, 5.7, 2.3, 4.9, 2.0, 6.7, 2.0, 4.9, 1.8, 5.7, 2.1,
            6.0, 1.8, 4.8, 1.8, 4.9, 1.8, 5.6, 2.1, 5.8, 1.6, 6.1, 1.9, 6.4, 2.0, 5.6, 2.2, 5.1, 1.5, 5.6, 1.4, 6.1, 2.3, 5.6, 2.4, 5.5,
            1.8, 4.8, 1.8, 5.4, 2.1, 5.6, 2.4, 5.1, 2.3, 5.1, 1.9, 5.9, 2.3, 5.7, 2.5, 5.2, 2.3, 5.0, 1.9, 5.2, 2.0, 5.4, 2.3, 5.1, 1.8,
        ];

        let kmean: KMeans<f32, 8, _> = KMeans::new(&samples, 150, 2, EuclideanDistance);
        let rnd = rand::rngs::StdRng::seed_from_u64(3);
        let conf = KMeansConfig::build()
            .random_generator(rnd)
            .abort_strategy(AbortStrategy::NoImprovementForXIterations {
                x: 5,
                threshold: 0.0005f32,
                abort_on_negative: true,
            })
            .build();
        let res = kmean.kmeans_minibatch(30, 3, 100, KMeans::init_kmeanplusplus, &conf);

        // SHOULD solution
        let should = KMeansShouldResult {
            distsum: 31.751724,
            sample_dims: 2,
            assignments: vec![
                1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
                1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 2, 0, 0, 0, 0,
                0, 2, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 2, 2, 2, 2, 2, 2, 0, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 0, 2, 2, 2,
                2, 2, 2, 0, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 0, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
            ],
            centroid_distances: vec![
                0.01267357, 0.01267357, 0.037013218, 0.0083339345, 0.01267357, 0.06493769, 0.005315068, 0.0083339345, 0.01267357,
                0.03569244, 0.0083339345, 0.023994308, 0.040032074, 0.17305095, 0.08135281, 0.01361693, 0.042296212, 0.005315068,
                0.052296188, 0.0009754318, 0.05965469, 0.01361693, 0.2300321, 0.09757918, 0.19097538, 0.023994308, 0.029277302,
                0.0083339345, 0.01267357, 0.023994308, 0.023994308, 0.01361693, 0.03569244, 0.01267357, 0.0083339345, 0.08135281,
                0.037013218, 0.040032074, 0.037013218, 0.0083339345, 0.029654715, 0.029654715, 0.037013218, 0.114560306, 0.19625838,
                0.005315068, 0.023994308, 0.01267357, 0.0083339345, 0.01267357, 0.16893126, 0.075018674, 0.3967574, 0.08893235, 0.12545326,
                0.041105583, 0.24284437, 1.0950203, 0.09154017, 0.16545418, 0.73588943, 0.043714777, 0.1880627, 0.16893126, 0.48719388,
                0.0176275, 0.075018674, 0.13849738, 0.075018674, 0.20458452, 0.48719242, 0.08893235, 0.3967574, 0.17501816, 0.00023629011,
                0.0176275, 0.25936627, 0.39401728, 0.075018674, 0.73588943, 0.29415, 0.4567587, 0.17154106, 0.3755859, 0.075018674,
                0.12197524, 0.1958878, 0.010670955, 0.039367035, 0.08893235, 0.023714393, 0.09849672, 0.10197579, 1.0950203, 0.009801679,
                0.02284512, 0.009801679, 0.00023629011, 1.7306727, 0.039367035, 0.4355868, 0.20323333, 0.1412727, 0.061272323, 0.09970424,
                1.1500959, 0.18893182, 0.6500962, 0.12950772, 0.5397043, 0.18578245, 0.07146844, 0.0048020603, 0.28166473, 0.31597906,
                0.12166494, 0.057154696, 1.3967625, 1.9475476, 0.5271919, 0.09813553, 0.39754698, 1.3716642, 0.45244873, 0.03303728,
                0.27774292, 0.48719242, 0.45244873, 0.008919688, 0.26440942, 0.34440956, 0.7593119, 0.03146885, 0.4730368, 0.4110758,
                0.39460605, 0.13656716, 0.057154696, 0.48719242, 0.020684395, 0.13656716, 0.25342983, 0.20323333, 0.20637095, 0.24323381,
                0.17754751, 0.2991156, 0.10990014, 0.08578265, 0.24068421,
            ],
            centroids: vec![4.297827, 1.3152173, 1.4716982, 0.28679252, 5.529412, 2.0372543],
        };

        assert_kmeans_result_eq(should, res);
    }
}