Decision Workspace
trust-region-least-squares vs scirs2-optimize vs cartan-optim
Side-by-side comparison of Rust crates
58
trust-region-least-squares
experimentalv0.11.0
A dense trust-region-reflective nonlinear least-squares solver that reproduces SciPy's least_squares (method='trf') bit-for-bit on its unbounded path for arbitrary dimension n, every loss (linear/soft_l1/huber/cauchy/arctan), and 2-point Jacobians, with an injectable host-runtime numerics backend for exact SVD, BLAS, and power dispatch.
52
scirs2-optimize
growingv0.6.5
Optimization module for SciRS2 (scirs2-optimize)
60
cartan-optim
experimentalv0.10.0
Riemannian optimization algorithms for cartan: Riemannian gradient descent, conjugate gradient, trust region
Core Metrics
| trust-region-least-squares | scirs2-optimize | cartan-optim | |
|---|---|---|---|
| Health Score | 58 | 52 | 60 |
| Total Downloads | 3.7K | 40.4K | 789 |
| 30d Downloads | 0 | 0 | 0 |
| Dependents | 61 | 374 | 21 |
| Releases | 5 | 39 | 19 |
| Last Updated | 40d ago | 70d ago | 26d ago |
| Age | 3m | 1y 6m | 7m |
Health Breakdown
trust-region-least-squares
Maintenance
17
Quality
13
Community
11
Popularity
4
Documentation
13
scirs2-optimize
Maintenance
15
Quality
8
Community
13
Popularity
6
Documentation
10
cartan-optim
Maintenance
17
Quality
15
Community
10
Popularity
3
Documentation
15
Technical Details
| trust-region-least-squares | scirs2-optimize | cartan-optim | |
|---|---|---|---|
| Version | 0.11.0 | 0.6.5 | 0.10.0 |
| Stable (≥1.0) | ✗ No | ✗ No | ✗ No |
| License | MIT | Apache-2.0 | MIT |
| Dependencies | 5 | 17 | 7 |
| Crate Size | 634KB | 1.4MB | 17KB |
| Features | 2 | 7 | 3 |
| Yanked % | 0.0% | 25.6% | 0.0% |
| Edition | 2021 | 2021 | 2024 |
| MSRV | — | — | 1.89 |
| Owners | 1 | 1 | 1 |
Links
Quick Verdict
- •cartan-optim leads with a health score of 60/100, but none of the options score above 80.
- •scirs2-optimize has the most downloads (40.4K), suggesting wider adoption.
- •scirs2-optimize is depended on by 374 crates — strongest ecosystem trust.