Improve visualization to distinguish repeated from single co-participation

[timofruehwirth]

The current spring layout treats all edges equally for positioning, causing single-event co-participants to cluster together. This visual proximity invites misinterpretation as social closeness.

Problem:

• Spatial proximity in network visualizations is naturally read as relational proximity
• Single co-attendance at a large event should not visually cluster people together
• The current visualization does not distinguish repeated from incidental co-participation

Proposed solution:

Adjust layout and styling to reflect a meaningful data distinction:

• Layout: Transform edge weights so single co-attendance (weight = 1) exerts minimal pull; repeated co-participation (weight ≥ 2) positions nodes closer
• Edge styling: Single co-attendance as faint dashed lines; repeated co-participation as solid lines with width scaled to weight

Framing (important):

• This is not a claim that repeated co-participation equals genuine relationship
• This is a visual encoding that reflects a meaningful distinction in the data
• Dashed lines signal "interpret with caution," not "unimportant"

To do:

• Add weight transformation for layout calculation
• Separate edge drawing: single (dashed, faint) vs. repeated (solid, prominent)
• Add brief explanation in notebook of why visualization is structured this way

Reference: See md file for implementation.