NEExT 0.2.2

Unsupervised Graph Analysis Framework.

NEExT

Unsupervised Graph Analysis Framework

UGAF is a frameowork for analysis of a collection for graphs. This includes functionality such as:

• Cleansing and standardizing graph data.
• Creating node and structural embedding for nodes in the graph collection.
• Creating embedding for graphs (graph embedding).

Instalation Process

UGAF uses Python 3.x (currently tested using Python 3.11). You can install UGAF using the following:

pip install git+https://${GIT_USERNAME}:${GIT_PASSWORD}@github.com/elmspace/ugaf.git

where GIT_USERNAME and GIT_PASSWORD are environment variables, which you can set in terminal by using:

export GIT_USERNAME=<your git username>

and

export GIT_PASSWORD=<your git classic token>

** Future version of UGAF will be on public PyPi, which would allow you to install it using pip directly.

Graph Data Format

Graph Collection Data

Here, we cover data format for creating graph collection data: UGAF expects input graph data to be in csv format. To create a graph collection, you would need two csv files: The edge csv file, contains the relationship between nodes (how they are connected). Here is an example:

node_a	node_b
1	2
3	2
.	.

The node graph mapping csv file contains the relationship between nodes and graphs. In other words, which node belongs to which graph. Here is an example:

node_id	graph_id
0	1
1	1
2	1
3	2
4	2
.	.

Here, nodes (0, 1 and 2) belong to graph 1 and nodes (3 and 4) belong to graph 2. ** Note that UGAF does not make the assumption that each graph is singhle connected component. Although you could filter for only connected component, as you shall in the example section.