JG.

Learning Social Network Analysis in R

This week's module we work with social network analysis in R. At first this type of thing looks pretty weird and not really sure how it would be useful in analysis but then I realized how it shows relations. I fired up R studio and tried the code provided

library(GGally)

library(network)

library(sna)

library(ggplot2)

net = rgraph(10, mode = "graph", tprob = 0.5)

net = network(net, directed = FALSE)

network.vertex.names(net) = letters[1:10]

ggnet2(net)

ggnet2(net, node.size = 6, node.color = "black", edge.size = 1, edge.color = "grey")

Screenshot 2024-04-08 at 2.04.57 AM.png


while that creates this cool little visual it doesn't really tell me much. Its not very useable like this so I now need to try to implement this with some data. As I am always trying to find ways to make senes out of my collected Zed Run data I decided to try to import a dataframe of my stable and find ways to build relations.

This was incredibly challenging as I learned I had to make a matrix to show similarities.The easiest would of been to simply connect horses by bloodline or simple characteristics. I decided to group horses by winrate with a threashold of 1.

Screenshot 2024-04-08 at 1.41.46 AM.png

The result gave ma few horses on islands by themselves. Here is the dataframe to help you understand the visualization sorted by WinRate.

HorseID

Name

WinRate

Bloodline

HorseType

517208

Sphinxx

19.69

Finney

Mare

591430

Grand Jefe

15.13

Buterin

Colt

22295

Red Light Green Light

14.81

Buterin

Mare

589053

Bondage

14.63

Buterin

Stallion

521223

Same Gamble

14.29

Buterin

Colt

602693

ManeSidePiece

14.06

Buterin

Colt

400924

Tyr the Brave

12.69

Nakamoto

Stallion

498085

Protonic

12.66

Szabo

Mare

132975

Breast In Breed

11.5

Buterin

Mare

505039

Pleasurable

10.53

Nakamoto

Mare

436742

No Donkey Left Behind

10.07

Szabo

Filly

14751

Limited Tier

9.03

Buterin

Stallion

133962

C h e e r i o  !

8.47

Szabo

Mare

14435

Trusting Starlight

7.97

Buterin

Mare

600571

E I E I O

6.76

Szabo

Colt

76556

Anime Princess

6.67

Szabo

Mare

442758

Irrational Spending

5.36

Buterin

Filly

436743

Cousin Luke

5.32

Buterin

Colt

601535

FailurIsNotAnOption

4.5

Buterin

Colt

605743

How Are My Brake Lights

4.35

Buterin

Filly

392732

Diamonds Edge

3.96

Szabo

Mare

149466

Best In Snow

3.25

Buterin

Colt

604467

Donkey In Chains

2.63

Buterin

Filly

603066

Junkyard Donkey

2.44

Buterin

Colt

452140

2jzgte

1.92

Buterin

Filly

29968

Petergate

0

Buterin

Stallion

As you see Petergate has 0% win rate so it is on his own island as well as Sphinxx who is almost 5% above the next horse.

After more research I realize that this is not the best use or example of social network analysis. Finding datasets were objects may have multiple relations with different strengths of relations would be a better use of this. However for simply showing realtionships this is was a fine example. I had the option of using Excel to show this how ever I rather work on my R skills or python skills versus doing these visulizations in R.