Metadata-Version: 2.4
Name: renewable-huber-native-cpu
Version: 0.7.0
Classifier: Development Status :: 2 - Pre-Alpha
Classifier: Programming Language :: Python :: 3
Classifier: Programming Language :: Python :: 3.10
Classifier: Programming Language :: Python :: 3.11
Classifier: Programming Language :: Python :: 3.12
Classifier: Programming Language :: Python :: 3.13
Classifier: Programming Language :: Rust
Classifier: Operating System :: Microsoft :: Windows
Classifier: Operating System :: POSIX :: Linux
Classifier: Operating System :: MacOS
Requires-Dist: renewable-huber==0.7.0
License-File: LICENSE
License-File: NOTICE
Summary: Optional Rust CPU engine for renewable-huber.
Author: Renewable Huber contributors
License-Expression: Apache-2.0
Requires-Python: >=3.10, <3.14
Description-Content-Type: text/markdown; charset=UTF-8; variant=GFM
Project-URL: Documentation, https://funtrollor.github.io/renewable-huber/native-core-p1/
Project-URL: Issues, https://github.com/Funtrollor/renewable-huber/issues
Project-URL: Source, https://github.com/Funtrollor/renewable-huber

# renewable-huber-native-cpu

Optional Rust/Rayon CPU engine for
[`renewable-huber`](https://github.com/Funtrollor/renewable-huber). Version
0.7.0 requires exactly `renewable-huber==0.7.0`; the base package owns the
public estimator API, validation and portable checkpoint format.

```bash
python -m pip install renewable-huber-native-cpu==0.7.0
```

Release wheels support CPython 3.10–3.13 on Windows x86-64, manylinux2014
x86-64/aarch64 and macOS x86-64/Apple Silicon. Installing a
wheel does not require Rust.

```python
from renewable_huber import RenewableHuberRegressor

model = RenewableHuberRegressor(backend="native_cpu", n_jobs=-1)
model.fit(X, y)
```

`n_jobs=None` uses the shared Rayon pool, `-1` uses all available logical CPUs,
and a positive integer creates an estimator-local pool of that size. When an
outer joblib or GridSearchCV layer is parallel, use `n_jobs=1` to avoid nested
oversubscription.

CPU `backend="auto"` may select this engine for a sufficiently large workload
when a bounded, host-local runtime calibration demonstrates a conservative
advantage. It never reads a CPU model string or persists a machine-specific
crossover map. Explicit `backend="native_cpu"` never silently falls back.

Inputs are dense float32/float64 host arrays and predictions are NumPy arrays.
See the project [support matrix](https://github.com/Funtrollor/renewable-huber/blob/main/docs/support-matrix.md)
for the complete contract.

