🙏 Feature request: funtion to import nodes.csv & edges.csv (aka. " Gephi Import CSV Data")

[adriens]

❔ About

In various graph dataviz tools (like Gephi) it is very common to import files to create a graph with csv:

• nodes.csv
• edges.csv

The main idea would to make it easy to do the same on yFiles.

💡 Feature request

• Implement the same feature as Import CSV Data in yFiles (with the exact same requirement)
• Add it to the examples

⏳ Workaround

As a workaround I'm using the graphml import feature

[yGuy]

That might work:

import pandas as pd

# Read CSV files
nodes_df = pd.read_csv('nodes.csv')
edges_df = pd.read_csv('edges.csv')

# Create a new graph widget
w = GraphWidget()

# Add nodes with all attributes
w.nodes = []
for _, row in nodes_df.iterrows():
    node_attrs = row.to_dict()
    node_id = node_attrs.pop('id')  # Assuming 'id' is the column for node identifiers
    w.nodes.append({
        "id": node_id,
        "properties": node_attrs
    })

# Add edges with all attributes
w.edges = []
for _, row in edges_df.iterrows():
    edge_attrs = row.to_dict()
    edge_id = edge_attrs.pop('id', None)  # Assuming 'id' is the column for edge identifiers
    source = edge_attrs.pop('source')  # Assuming 'source' is the column for edge source
    target = edge_attrs.pop('target')  # Assuming 'target' is the column for edge target
    w.edges.append({
        "id": edge_id if edge_id is not None else f"{source}-{target}",
        "start": source,
        "end": target,
        "properties": edge_attrs
    })

# Set the graph as directed (you can change this if your graph is undirected)
w.directed = True

# Display the graph
display(w)

[adriens]

Thanks a lot for the code. I'll give it a try very soon.
I'll release a content on yWorks/yfiles-jupyter-graphs next week, with the graphml import, including a blog post and a video