![Animation of the formation of RFs during aSSC learning.] (http://invibe.net/cgi-bin/index.cgi/SparseHebbianLearning?action=AttachFile&do=get&target=ssc.gif)

• (!) tl;dr : [Download the code](https://github.com/meduz/shl_scripts/archive/master.zip). Or directly from the command-line, do

` wget https://github.com/meduz/shl_scripts/archive/master.zip unzip master.zip -d shl_scripts cd shl_scripts/ ipython notebook`

• This is a collection of python scripts to test learning strategies to efficiently code natural image patches. This is here restricted to the framework of the [SparseNet algorithm from Bruno Olshausen](http://redwood.berkeley.edu/bruno/sparsenet/).
• this has been published as Perrinet, Neural Computation (2010) (see http://invibe.net/LaurentPerrinet/Publications/Perrinet10shl ):

`bibtex @article{Perrinet10shl, Author = {Perrinet, Laurent U.}, Doi = {10.1162/neco.2010.05-08-795}, Journal = {Neural Computation}, Keywords = {Neural population coding, Unsupervised learning, Statistics of natural images, Simple cell receptive fields, Sparse Hebbian Learning, Adaptive Matching Pursuit, Cooperative Homeostasis, Competition-Optimized Matching Pursuit}, Month = {July}, Number = {7}, Title = {Role of homeostasis in learning sparse representations}, Url = {http://invibe.net/LaurentPerrinet/Publications/Perrinet10shl}, Volume = {22}, Year = {2010}, Annote = {Posted Online March 17, 2010.}, }`

• all comments and bug corrections should be submitted to Laurent Perrinet at Laurent.Perrinet@gmail.com
• find out updates on http://invibe.net/LaurentPerrinet/SparseHebbianLearning

sh python setup.py clean build install

This piece of code is distributed under the terms of the GNU General Public License (GPL), check http://www.gnu.org/copyleft/gpl.html if you have not red the term of the license yet.

Be sure to have :

• a computer (tested on Mac, Linux) with python + numpy (on macosx, you may consider using [HomeBrew](https://github.com/meduz/dotfiles/blob/master/init/osx_brew_python.sh),
• grab the sources from the [Download the code](https://github.com/meduz/shl_scripts/archive/master.zip),
• These scripts should be platform independent, however, there is a heavy bias toward unix users when generating figures.

• README.md : this file
• learn.py : the scripts (see Contents.m for a script pointing to the different experiments)
• ssc.py : the individual experiments
• IMAGES_sparsenet.mat : the image files (if absent, they get automagically downloaded from [this link](http://invibe.net/LaurentPerrinet/SparseHebbianLearning?action=AttachFile|this page).

• matlab_code : some obsolete matlab code

* get the code with CLI wget https://github.com/meduz/shl_scripts/archive/master.zip. * decompress unzip master.zip -d shl_scripts * get to the code cd shl_scripts

* run the main script python learn.py

* remove SSC related files to start over rm -f IMAGES_*.mat.pdf *.hdf5

* 2.0 - 2015-05-07:

• use sklearn to do all the hard low-level work, in particular sklearn.decomposition.SparseCoder see http://scikit-learn.org/stable/auto_examples/decomposition/plot_image_denoising.htm and http://www.cs.utexas.edu/~leif/pubs/20130626-scipy-johnson.pdf
• transform to a class to just do the Sparse Hebbian Learning (high-level) experiments (getting data from an image folder, learning, coding, analyszing)

* 1.1 - 2014-06-18:

• documentation
• dropped Matlab support

• 1.0 - 2011-10-27 : initial release