hurwitzlab/imicrobe-mash-node2vec

Name: imicrobe-mash-node2vec

Owner: Hurwitz Lab

Description: Scripts to learn and cluster a vector-space embedding on all-vs-all iMicrobe MASH distance matrix.

Created: 2017-11-05 15:00:41.0

Updated: 2017-11-06 02:17:17.0

Pushed: 2017-11-06 02:39:40.0

Homepage: null

Size: 13

Language: Shell

GitHub Committers

User	Most Recent Commit	# Commits

Other Committers

User	Email	Most Recent Commit	# Commits

README

imicrobe-mash-node2vec

Scripts to learn and cluster a vector-space embedding on all-vs-all iMicrobe MASH distance matrix.

History

The first version of these scripts used the node2vec implementation in SNAP. However that implementation does not work. The reference implementation works and is efficient thanks to the excellent word2vec implementation in gensim.

Requirements

• ~SNAP~
• Python 3.6+
• Hurwitz Lab node2vec fork

Install

Create and activate a Python 3.6+ virtual environment.

thon3.6 -m venv ~/n2v
urce activate ~/n2v/bin/activate
) $ pip install -r requirements.txt

or

nda create -n n2v gensim networkx
urce activate n2v
) $

Install Hurwitz Lab fork of node2vec.

) $ git clone https://github.com/hurwitzlab/node2vec.git
) $ cd node2vec
) $ pip install -e .

Install on Stampede2

The node2vec algorithm is computationally expensive. To install an efficient implementation on Stampede2 it is important to use, as much as possible, the native optimized BLAS library.