Metadata-Version: 2.4
Name: tapkee
Version: 1.3.1
Summary: C++ library for dimensionality reduction
Author-Email: Sergey Lisitsyn <lisitsyn@hey.com>, =?utf-8?q?Fernando_Jos=C3=A9?= <fernando.iglesiasg@gmail.com>
License-Expression: BSD-3-Clause
Classifier: Development Status :: 4 - Beta
Classifier: Intended Audience :: Science/Research
Classifier: Programming Language :: C++
Classifier: Programming Language :: Python :: 3
Classifier: Topic :: Scientific/Engineering
Project-URL: Homepage, https://tapkee.lisitsyn.me
Project-URL: Repository, https://github.com/lisitsyn/tapkee
Project-URL: Issues, https://github.com/lisitsyn/tapkee/issues
Requires-Python: >=3.11
Requires-Dist: numpy
Description-Content-Type: text/markdown

# Tapkee

Python bindings for the Tapkee C++ dimensionality reduction library.

Tapkee provides efficient implementations of over 20 dimensionality reduction methods
with a focus on spectral techniques. It leverages Eigen for linear algebra and ARPACK
for large-scale eigenproblems.

## Installation

```bash
pip install tapkee
```

## Usage

```python
import numpy as np
import tapkee

data = np.random.randn(3, 100)  # 3 features, 100 samples
embedding = tapkee.embed(data, method="isomap", num_neighbors=15)
```

## Supported methods

- **Spectral:** Isomap, Landmark Isomap, Diffusion Maps, Laplacian Eigenmaps
- **Local:** LLE, Kernel LLE, Hessian LLE, LTSA, Linear LTSA, LPP, NPE
- **Linear:** PCA, Kernel PCA, Factor Analysis, Random Projection
- **Modern:** t-SNE, Barnes-Hut-SNE, Stochastic Proximity Embedding
- **Other:** MDS, Landmark MDS, Manifold Sculpting

## Links

- [Documentation](https://tapkee.lisitsyn.me)
- [Source code](https://github.com/lisitsyn/tapkee)
- [Issue tracker](https://github.com/lisitsyn/tapkee/issues)

## Citation

If you use Tapkee in a publication, please cite:

> Sergey Lisitsyn, Christian Widmer, Fernando J. Iglesias Garcia.
> Tapkee: An Efficient Dimension Reduction Library.
> Journal of Machine Learning Research, 14: 2355-2359, 2013.

## License

BSD 3-Clause
