Top pick: Lรณpez de Prado cMDA + HRP (mlfinlab / PyPortfolioOpt (mixed)). Fills the gap: Portfolio-level orthogonality โ which the production baseline (Pearson + Spearman + h_norm + Tier-2 MI-vs-target) cannot detect.
Lane = how usable the metric is today. Top lane = production-baseline duplicates (Pearson, Spearman). Below = permissive FOSS / GPL / unmaintained / code-released-but-not-packaged / no-FOSS. Per-circle text = first 3 letters of language (Pyt / R / Jav).
| Metric | Library | License | Lang | Maint | Complexity | Param-free? | URL |
|---|---|---|---|---|---|---|---|
| Clustered MDA (cMDA) + ONC | mlfinlab (mixed) / emoen/Machine-Learning-for-Asset-Managers (open) | mixed / open | Python | โ | O(N^2) clustering + O(N T trees) | no | โ |
| HRP โ Riskfolio | Riskfolio-Lib | BSD | Python | โ | ? | โ | |
| Hierarchical Risk Parity (HRP) | PyPortfolioOpt | MIT | Python | โ | O(N^2 log N) | yes | โ |
| ICA portfolio optimization | FastICA (sklearn) + custom opt | Apache-2 | Python | โ | ? | โ | |
| LOBFrame | Open-source per paper | open | Python | โ | ? | โ | |
| Alpha101 bias-correction | no public code | n/a | n/a | โ ๏ธ | ? | โ | |
| AlphaAgent | no public code | n/a | n/a | โ ๏ธ | ? | โ | |
| WorldQuant BRAIN regression_neut/vector_neut | platform-only | n/a | WQ-Expression-language | โ | ? | โ |