Metadata-Version: 2.1
Name: ksu
Version: 0.2.2
Summary: Implementation of the KSU compression algorithm https://www.cs.bgu.ac.il/~karyeh/compression-arxiv.pdf
Home-page: https://github.com/nimroha/ksu_classifier
Author: Nimrod Morag, Yuval Nissan
Author-email: nimrod.morag@gmail.com
License: MIT
Platform: UNKNOWN
Classifier: License :: OSI Approved :: MIT License
Classifier: Programming Language :: Python
Classifier: Programming Language :: Python :: 2
Classifier: Programming Language :: Python :: 2.7
Classifier: Development Status :: 4 - Beta
Requires-Python: >=2.7.0
Description-Content-Type: text/markdown
Requires-Dist: numpy
Requires-Dist: sklearn
Requires-Dist: scipy
Requires-Dist: psutil
Requires-Dist: editdistance
Requires-Dist: mnist
Requires-Dist: tqdm
Requires-Dist: requests-toolbelt

## KSU Compression Algorithm Implementation ##

Algortihm 1 from [Nearest-Neighbor Sample Compression: Efficiency, Consistency, Infinite Dimensions](https://arxiv.org/abs/1705.08184)

Installation
------------
* With pip: `pip install ksu`
* From source:
    * `git clone --recursive https://github.com/nimroha/ksu_classifier.git`
    * `cd ksu_classifier`
    * `python setup.py install`

 Usage
 -----
 This package provides a class `KSU(Xs, Ys, metric, [gram, prune, logLevel, n_jobs])`

 `Xs` and `Ys` are the data points and their respective labels as [numpy  arrays](https://docs.scipy.org/doc/numpy/reference/generated/numpy.array.html) 

 `metric` is either a callable to compute the metric or a string that names one of our provided metrics (print `ksu.KSU.METRICS.keys()` for the full list) 

 `gram` _(optional, default=None)_ a precomputed [gramian matrix](http://mathworld.wolfram.com/GramMatrix.html), will be calculated if not provided.

 `prune` _(optional, default=False)_ a boolean indicating whether to prune the compressed set or not (Algorithm 2 from [Near-optimal sample compression for nearest neighbors](https://arxiv.org/abs/1404.3368))

 `logLevel _(optional, default='CRITICAL')_ a string indicating the logging level (set to 'INFO' or 'DEBUG' to get more information)

 `n_jobs` _(optional, default=1)_ an integer defining how many cpus to use, pass -1 to use all. For n_jobs below -1, (n_cpus + 1 + n_jobs) are used. Thus for n_jobs = -2, all CPUs but one are used.

  <br>

  `KSU` provides a method `compressData([delta])`

  Which selects the subset with the lowest estimated error with confidence `1 - delta`.

  You can then run `getClassifier()` which returns a 1-NN Classifer (based on [sklearn's K-NN](http://scikit-learn.org/stable/modules/generated/sklearn.neighbors.KNeighborsClassifier.html)) fitted to the compressed data.

  Or, run `getCompressedSet()` to get the compressed data as a tuple of numpy arrays `(compressedXs, compressedYs)`.

  <br>

  See `scripts/` for example usage

