"friends are just like skittles" - dancingapple
“友谊就像彩虹糖” - dancingapple
其实太阳系最牛的R搞这些易如反掌，只是数据源有点不靠谱，略表失望。所以只是意思一下，点到即止。这次需要用到的是三个R库，“ jsonlite ”，解析Json格式用的；“igraph”，画一些基本网络图的；“rgl”，画一些3D图的。
这次笔记的目的是优美优雅地可视化那些甜蜜的友谊。
It's so easy to manipulate data in the almighty R. Nevertheless, I was a little disappointed by the data source, which I did some fact-checking and found not so update-to-date or accurate. The purpose of this note is to show some proof-of-concept, not intended to serve as thorough analysis anyway.
We'll mainly use three R libraries, "jsonlite" for parsing Json formatted data, "igraph" for denoting and plotting some basic network/graph, and "rgl" for advanced 3D visualization.
The whole purpose is to elegantly present the sweet friendship in a colorful way.
1. 清内存，读库 | clear memory and load libraries
rm(list = ls())
require(jsonlite)
require(igraph)
require(rgl)
2. 设定一个简单函数 | create a simple function
读取你的粉丝和你粉的丝，并转化为一个网络图
load your followers or the people you follow, and turn the relationship into a graph
ID.network <- function(user.ID){
#核心webapi | the core webapi access operation
fan.res <- fromJSON(paste("https://api.steemdata.com/busy.org/",
user.ID,"/followers", sep = ""))
follow.res <- fromJSON(paste("https://api.steemdata.com/busy.org/",
user.ID,"/following", sep = ""))
N.fan <- nrow(fan.res)
N.follow <- nrow(follow.res)
unique.name <- union(follow.res$name, c(fan.res$name,user.ID))
N.name <- length(unique.name)
mtx <- matrix(0,N.name,N.name)
colnames(mtx) <- rownames(mtx) <- unique.name
mtx[user.ID, follow.res$name] <- 1
mtx[fan.res$name, user.ID] <- 1
tmp.g <- graph_from_adjacency_matrix( mtx )
return(tmp.g)
}
3. 请出 
练手 | let's try an example
# 替换justyy为你想偷窥的ID
# feel free to replace "justyy" as any steemit ID you are curious
g1 <- ID.network(user.ID = "justyy")
plot(g1)
哗～这就是他的朋友圈。。。但是有点挤，色泽单调
wow~ that's his friendship... a little bit crowded, and dry in terms of color
好吧，来点颜色，比如256种彩虹色。并想象大家只是无名的小花。
ok. let's paint some colors, e.g. a series of rainbow colors. Also, let's imagine friends are all anonymous flowers.
plot(g1, vertex.color = rainbow(256), vertex.label=NA)
太平，太飞机场了，通过rgl库的3D可视化，搞点凸凹有致～
it's still flat. let's apply 3D visualization function in the rgl library:
'rglplot(g1, vertex.color = rainbow(256), vertex.label=NA)'
好了，可远观，可放大缩小旋转亵玩焉。。。
done! then you can zoom-in & -out, rotate and play with it from all different angles
4. 搞个循环多挖点人？ | make a loop to include more users
# 你可以直接定义想挖的人，把它们组织成一个字符串数组
# feel free to define a group of users you are curious about, and make them as a vector of strings
user.ID.vec <- c("justyy","susanli3769", "tumutanzi", "dapeng", "tvb", "bigyellow","jubi")
g.union <- ID.network(user.ID = user.ID.vec[1])
for(k in 2 : length(user.ID.vec)){
g2 <- ID.network(user.ID = user.ID.vec[k])
g.union <- union(g.union, g2)
}
如果他们是红花，大伙儿是绿叶
let's pretend they are red flowers and the rest are green leaves.
plot.igraph(g.union, layout=layout_with_fr, vertex.label=NA, edge.arrow.size = 0.1,
vertex.color = ifelse(is.element(V(g.union)$name,user.ID.vec),
yes = "red", no = "green"),
vertex.size = ifelse(is.element(V(g.union)$name,user.ID.vec),
yes = 10, no = 3))
或者大家都是无名的彩色小花儿
又或者，大家都是亲密无间的甜蜜的彩虹糖
or, all of us are just colorful flowers
or, friends are just like sweet sweet skittles
plot.igraph(g.union, layout=layout_with_fr,vertex.size=6,edge.arrow.size = 0.1,
vertex.label=NA, vertex.color = rainbow(256), edge.color = "green")
rglplot(g.union, layout=layout_with_fr,
vertex.color = rainbow(256), vertex.label=NA)
5. 打完，收工 | Done!
。。。截止到现在目前为止，我们都在干什么呢？其实我们在创作艺术，用太阳系最牛的R
... wait a min, what are we doing so far? we are actually creating some art-works, using almighty R
又及，如果想再深挖的话，igraph有很多很牛的网络分析函数，比如Betweenness centrality（中文都不知道咋翻）。有心人可以细玩。
BTW, there are tons of cool graph analytical functions in the "igraph" library, to compute key statistics of the network/graph. Enjoy if interested.