WHAT:
-------------------------------------------
* Implementations of fast exact k-means algorithms as described in http://arxiv.org/abs/1602.02514 and implementations of turbo-charged mini-batch k-means as described in http://arxiv.org/pdf/1602.02934

for interfaces
(LIB) Shared library with accompanying C++ header file
(EX) Command-line exectuble
(PY) Python library


REQUIREMENTS:
-------------------------------------------
Minimal installation requirements:
-- C++ compiler supporting C++11
-- Linux operating system

Optional but recommended:
-- BLAS implementation, we recommend this one : http://www.openblas.net/

Specific to Python library:
-- Python and Cython


CONFIGURATION:
-------------------------------------------
In Makefile, set USEBLAS to either NO or YES
if USEBLAS = YES, then set LIBBLASDIR, INCBLASDIR (unless blas paths will be found automatically)


BUILDING:
-------------------------------------------
For (LIB) and (EX) and (PY) : `make all'
For (EX) : `make main'
For (LIB) : `make lib'

USING: 
-------------------------------------------
(EX) If succesfully installed, you should find an executable in directory bin
Run the executable with -h flag to see the options

(LIB) You need to add lib directory to your LD_LIBRARY_PATH : put the following line in your ~/.bashrc file
export LD_LIBRARY_PATH=${LD_LIBRARY_PATH}:/path/to/kmeans/lib
 
(PY) If successfully installed, you should be able to `import kmeans' when in directory lib. 
To use from a different directory, 

(a) as per (LIB), and

(b) add the path to lib to your python path, either by
export PYTHONPATH=${PYTHONPATH}:/path/to/kmeans/lib
or directly in your python script : import sys, sys.path.insert(0,'/path/to/kmeans/lib')
Example use is found in examples/examples.py



DOESN'T WORK?
-------------------------------------------
Please contact me at jnewling@idiap.ch