Overview

Repository for a tutorial at THW, Berkeley on parallel programming in Python.

This short tutorial introduces parallel programming with mpi4py and concurrent.futures, and applies it to recognition of hand-written digits.

Running the tutorial

The tutorial is in the form of Jupyter notebooks. You can run these notebooks:

remotely on mybinder.org: to do so, click the above badge.
locally on your computer. To do so, install Anaconda and install the requirements by typing

conda config --add channels conda-forge 
conda env create -n parallel_python -f install/environment.yml
source activate parallel_python

Name	Name	Last commit message	Last commit date
Latest commit RemiLehe Revert "Update Dockerfile" Oct 19, 2016 e88d907 · Oct 19, 2016 History 18 Commits
classification	classification	Progress in the serial implementation	Oct 15, 2016
data	data	Added the datasets	Oct 15, 2016
install	install	Added a Dockerfile	Oct 15, 2016
tutorial_images	tutorial_images	Additional modification of the notebooks	Oct 18, 2016
Classification_futures.ipynb	Classification_futures.ipynb	Additional modification of the notebooks	Oct 18, 2016
Classification_mpi4py.ipynb	Classification_mpi4py.ipynb	Additional modification of the notebooks	Oct 18, 2016
Classification_serial.ipynb	Classification_serial.ipynb	Added more explanations in the mpi4py code.	Oct 18, 2016
Dockerfile	Dockerfile	Revert "Update Dockerfile"	Oct 19, 2016
README.md	README.md	Update README.md	Oct 17, 2016
index.ipynb	index.ipynb	Added images in the index	Oct 17, 2016