##EkiNetworkX - Custom functions to compute networkX graph
========================================================
INSTALL
Install with pip:
pip install ekiNx
Example of usage:
>>> from ekiNx import core
>>> core.proclamer()
========================================================
This module provides complement functions for lib NetworkX
Function to compute bipartite graph and project to one set of nodes
FromDataFrame :
-
INPUT:
- data is a pandas dataframe
- u is the name of column for node 'projected'
- v is the name of column for node to project
- w is the name of column for weight
- alpha is a threshold to map or not a edge
-
OUTPUT:
- compute numpy array (dim 2 N)
- U: are nodes projected
- V: are node to project
- E: edges with weight attributes if weight columns filled
mapBipartite :
- create biGraph and map nodes/edges
- INPUT:
- U nodes 'projected'
- V nodes to project
- E: edges with weight attributes if weight columns filled
- OUTPUT:
- g: networkX biGraph
projectGraph :
- create biGraph and map nodes/edges
- INPUT:
- U biGraph a bigraph networkX object
- V is the set on
- weight_function to compute edges weight ( for instance jaccard)
- OUTPUT:
- g: networkX object representing a undirected graph
graphFromPandasAdjancyMatrix :
- map nodes and edges from adjancy matrix:
- INPUT:
- data is a pandas object representing an adjacency matrix
- OUTPUT:
- g: anetworkX object representing a undirected graph each node contain is label with column label
========================================================
Ce code est sous licence WTFPL.