surfiam-graph-vis

Python code to visualise preconfigured SRAM graphs, SRAM json exports of an organisation and getting statistics from such a json.

Dependencies

• Python 3.11 or higher
• networkx
• pyvis

Installation

Install a git branch as python package

pip install git+https://github.com/chStaiger/surfiam-graph-vis.git@<branch>

Checkout code and install python package

Installation with pip

```

HTTP

git clone https://github.com/chStaiger/surfiam-graph-vis.git

SSH

git clone git@github.com:chStaiger/surfiam-graph-vis.git

cd surfiam-graph-vis pip install -e . ```

Installation with uv

```

HTTP

git clone https://github.com/chStaiger/surfiam-graph-vis.git

SSH

git clone git@github.com:chStaiger/surfiam-graph-vis.git

cd surfiam-graph-vis uv build ```

Usage

To run the main script simply type in the terminal:

• Pip package

surfiamviz

• Uv package

uv run surfiamviz

To run the web tool use:

``` surfiamviz webtool

or

uv run surfiamviz webtool ```

Configuration

Standard graphs, node colours and edge colours can be submitted to the tool through a config file. We provide an example config file to illustrate how node and edge type determine the colour and to show two standard example graphs for an SRAM collaboration.

Nodes

A graph consists of nodes and edges. An edge defines a connection between two nodes.

In the section node_types you will find a list of all defined nodes for all possible SRAM graphs/networks. Each node has a name and a level e.g. ORG_ADMIN.name = "admin" and ORG_ADMIN.level.

The name groups nodes into functions or roles and gives them a common color as defined in the section [node_colors]. E.g. all nodes with the name admin will be plotted in green. You can change those colours in the section [node_colors]. If necessary you can also regroup those nodes by changing the labels and their respective counterpart in the colour section.

The node level determines on which level in the graph hierarchy the nodes of the respective type will be plotted. For SRAM the ORGANISATION is the top level indicated by the smallest number and appears in the graphs always on left side of the plot.

Edges

Similar to nodes we offer a section [edge_colors] to define the color of edges. All edge types are listed in this section too. You can extend that section. However, to add a new type of edge requires adjusting the code for the rendering of graphs coming from a real SRAM connection or json file.

[!CAUTION] The node types and the edge types are also used to render graphs from actual SRAM json exports. You can add edge types, node types and you can change the node type names or levels. Do not delete any of the existing node or edge types.

Configured SRAM graphs

Adding a new graph section

To define an SRAM graph start with a new section in the configuration toml-file:

[my_graph]

The name my_graph will be used in the commandline tool to determine which graph will be plotted:

surfiamviz graph -g my_graph

Adding new edge sets

In the new section you can set new edges and give the edges a type:

``` entities.edges = [["ORGANISATION", "COLLABORATION"], ["COLLABORATION", "APPGROUP"], ["COLLABORATION1", "CO_GROUP"] ]

entities.type = "BACKBONE" ``` Of course you can define several edge sets of different types.

The edge type, here "BACKBONE" is used in the section [edge_colors] to give all edges of the same type a color.

Colours

The names for colours are taken from the matplotlib colour scheme.

Plotting exported SRAM graphs

To visualise the actual situation of your SRAM organisation you can either directly plot the information from the server:

surfiamviz organisation -o test.html -c configs/sram_config.toml --server <SRAM SERVER> --token <SRAM TOKEN>

Or first download the information to a json file and plot it subsequently:

surfiamviz download --server <SRAM SERVER> --token <SRAM TOKEN --file sram_org.json surfiamviz organisation -o test.html -c configs/sram_config.toml --input sram_org.json

The file sram_org.json is used in the visualisation together with the configuration file to draw the graph:

We provide an example json file in data/sram_test_org.json.

The software will set the node types and the edge types. You can steer the colouring of nodes and edges in the configuration file in the section [node_colors] and [edge_colors].