Interactive comparison of Python, Rust, Rust-Parallel, and scikit-learn K-Means execution paths across CLI k-sweep runtime, throughput, sampled RSS, CPU/resource use, and clustering quality. Runtime is measured end-to-end, including CSV read/write and fitting k = 1..k_max.
| Implementation | Runtime median s | Runtime IQR s | RSS median MB | RSS MB / 1k | CPU s / 1k | Effective cores | Work units / s | ARI median |
|---|---|---|---|---|---|---|---|---|
| Rust - Parallel | 0.197 | 0.814 | 12.04 | 0.73 | 0.036 | 1.87 | 78,587,025 | 1.000 |
| Rust | 0.201 | 1.012 | 9.32 | 0.61 | 0.017 | 0.90 | 69,712,016 | 1.000 |
| Python | 0.806 | 2.274 | 104.93 | 7.41 | 0.082 | 0.91 | 15,538,872 | 1.000 |
| scikit-learn | 1.843 | 1.322 | 206.90 | 12.63 | 0.123 | 0.95 | 9,249,935 | 1.000 |