Metadata-Version: 2.4
Name: nirs4all-methods
Version: 1.0.3
Summary: nirs4all-methods — portable PLS / NIRS engine (full surface) over the libn4m C ABI.
Home-page: https://github.com/GBeurier/nirs4all-methods
Author: Grégory Beurier and contributors
License-Expression: CECILL-2.1
Project-URL: Homepage, https://github.com/GBeurier/nirs4all-methods
Project-URL: Repository, https://github.com/GBeurier/nirs4all-methods
Project-URL: Issues, https://github.com/GBeurier/nirs4all-methods/issues
Project-URL: Documentation, https://methods.nirs4all.org/
Requires-Python: >=3.10
Description-Content-Type: text/markdown
License-File: LICENSE
Requires-Dist: numpy>=1.21
Dynamic: license-file

# nirs4all-methods

nirs4all-methods is the full numerical/method engine distribution. The Python
distribution name is `nirs4all-methods`; the import package is `n4m`, matching
the stable `libn4m` C ABI and `n4m_*` exported symbols.

The wheel bundles the `libn4m` shared library, so `pip install nirs4all-methods`
is self-contained — no separate native build is required. The slim `pls4all`
distribution remains available for PLS-focused users; both distributions load
the same ABI family.

```python
import n4m
from n4m.estimators.regression.latent import PLS

print(n4m.version())      # project version + ABI, e.g. "1.0.3+abi.2.0.0"
print(n4m.abi_version())  # ABI triple, e.g. (2, 0, 0)

model = PLS(n_components=2).fit(X, y)
pred = model.predict(X)
```

`n4m` loads `libn4m` via `ctypes.CDLL` and exposes version/ABI helpers,
`Context`, and role-based packages such as `n4m.estimators`, `n4m.transform`,
`n4m.feature_selection`, `n4m.model_selection`, and `n4m.augmentation`.
Estimator-style wrappers live under those role packages; there is no flat
`n4m.sklearn` package in the ABI 2 surface.

Set `N4M_LIB_PATH` to point at a development `libn4m` build when overriding the
bundled wheel library.

See <https://github.com/GBeurier/nirs4all-methods> for documentation.
