On first glace, this seems like 2-case of k-discrete edge classes? Curious why they stopped at two, "colored" edges seem better for categorical distinct relationships than "positive/negative". I guess the authors intend complementary relationships, like the "love/fight" spectrum?

youtu.be/ADgS_vMGgzY

I'd like these to be
1. Stable, so I can render a frame-by-frame animation as days go by and relationships change
2. Readable, I wish I could work out how to encode distance as the "closeness" visually

Python. First, a Dict[str_name, List[Time_Interval[start,end]]] represents the time intervals a player plays.

Then, I turn this into a Matrix with M_ij := time that player_j plays while player_i is online. M is not symmetric!

Finally, I give M to networkx.spring_layout().

These are a few runs:

If we do any kind of consistency comparison, you would knock it out of the park. I don't know anyone who uploads so consistently, truly most qualified to state that Minecraft is your life

[Spoilers!]

What makes you say that? I'm curious to see how (if) Mark&Helly get out, learn more about Gemma from her POV, and see where Lumon goes with A World Without Pain (presumably other MDR depts had similar work?). Feels to me like important and mysterious questions remain

Can you link that video? I'm not finding it. Thanks!

Got log parsing to work reliably, started plotting the character network and found it's unstable (same parameters result in different visualization when rerun). Would like to see if your code is more consistent :)

Commenting to revisit for Whimsicraft Network visualization, once I fix the log parser. Thanks for consistently showing interesting network science, Not David!

What a deep cut :D
Excited to Hot Swap again sometime

This absolutely made my day, and possibly my week. Amazing! I love the dynamic pose, and even lava droplets from the belt-buckled bucket!

What happens when a pig is struck by lightning...? :think:

Prepping my Halloween skin...