Metadata-Version: 2.0
Name: hep-ml
Version: 0.5.0
Summary: Machine Learning for High Energy Physics
Home-page: https://github.com/arogozhnikov/hep_ml
Author: Alex Rogozhnikov
Author-email: axelr@yandex-team.ru
License: Apache 2.0
Keywords: machine learning,supervised learning,uncorrelated methods of machine learning,high energy physics,particle physics
Platform: UNKNOWN
Classifier: Intended Audience :: Science/Research
Classifier: License :: OSI Approved :: Apache Software License
Classifier: Programming Language :: Python :: 2.7 
Classifier: Programming Language :: Python :: 3.4 
Requires-Dist: numpy (>=1.9)
Requires-Dist: pandas (>=0.14.0)
Requires-Dist: scikit-learn (>=0.15.2)
Requires-Dist: scipy (>=0.15.0)
Requires-Dist: six
Requires-Dist: theano (>=0.7)

hep\_ml
=======

**hep\_ml** provides specific machine learning tools for purposes of
high energy physics (written in python).

.. figure:: https://github.com/arogozhnikov/hep_ml/blob/data/data_to_download/hep_ml_image.png
   :alt: hep\_ml, python library for high energy physics

   hep\_ml, python library for high energy physics

Main points
-----------

-  uniform classifiers - the classifiers with low correlation of
   predictions and mass (or some other variable(s))
-  **uBoost** optimized implementation inside
-  **UGradientBoosting** (with different losses, specially
   **FlatnessLoss** is very interesting)
-  measures of uniformity (see **hep\_ml.metrics**)
-  advanced losses for classification, regression and ranking for
   **UGradientBoosting** (see **hep\_ml.losses**).
-  **hep\_ml.nnet** - theano-based flexible neural networks
-  **hep\_ml.reweight** - reweighting multidimensional distributions
   (*multi* here means 2, 3, 5 and more dimensions - see GBReweighter!)
-  **hep\_ml.splot** - minimalistic sPlot-ting
-  **hep\_ml.speedup** - building models for fast classification (Bonsai
   BDT)
-  **sklearn**-compatibility of estimators.

Installation
------------

Basic installation:

.. code:: bash

    pip install hep_ml

If you're new to python and don't never used ``pip``, first install
scikit-learn `with these
instructions <http://scikit-learn.org/stable/install.html>`__.

To use latest development version, clone it and install with ``pip``:

.. code:: bash

    git clone https://github.com/arogozhnikov/hep_ml.git
    cd hep_ml
    sudo pip install .

Links
-----

-  `documentation <https://arogozhnikov.github.io/hep_ml/>`__
-  `notebooks, code
   examples <https://github.com/arogozhnikov/hep_ml/tree/master/notebooks>`__
-  `repository <https://github.com/arogozhnikov/hep_ml>`__
-  `issue tracker <https://github.com/arogozhnikov/hep_ml/issues>`__

Related projects
----------------

Libraries you'll require to make your life easier.

-  `IPython Notebook <http://ipython.org/notebook.html>`__ — web-shell
   for python
-  `scikit-learn <http://scikit-learn.org/>`__ — general-purpose library
   for machine learning in python
-  `yandex/REP <https://github.com/yandex/REP>`__ — python wrappers
   around different machine learning libraries (including TMVA) +
   goodies, required to plot learning curves and reports after
   classification. Required to execute *howto*\ s from this repository
-  `numpy <http://www.numpy.org/>`__ — 'MATLAB in python', vector
   operation in python. Use it you need to perform any number crunching.
-  `theano <http://deeplearning.net/software/theano/>`__ — optimized
   vector analytical math engine in python
-  `ROOT <https://root.cern.ch/>`__ — main data format in high energy
   physics
-  `root\_numpy <http://rootpy.github.io/root_numpy/>`__ — python
   library to deal with ROOT files (without pain)

License
-------

Apache 2.0, library is open-source.

Platforms
---------

Linux, Mac OS X and Windows are supported.

.. |travis status| image:: https://travis-ci.org/arogozhnikov/hep_ml.svg?branch=master
.. |Build status| image:: https://ci.appveyor.com/api/projects/status/kxatlw869t9ibbo3?svg=true
   :target: https://ci.appveyor.com/project/arogozhnikov/hep-ml
.. |PyPI version| image:: https://badge.fury.io/py/hep_ml.svg
   :target: http://badge.fury.io/py/hep_ml


