4회 R meetup 코드

작성일자 글쓴이

4회 R meetup은 동국대 통계학과 전산실에서 진행되었다. 약속대로 코드를 공유한다.

바로 공유를 했어야 되는데 개인적인 사정으로 인해 늦게 공유하게 되었는데 이점에 대해서 양해 부탁드린다.

 

먼저 ROAuth를 이용한 트위터 API 사용 권한 가져오는 예제이다.

키는 https://dev.twitter.com/apps/new 에서 등록하면 된다. 

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24

library(ROAuth)
library(twitteR)
cainfo <- system.file("CurlSSL", "cacert.pem", package = "RCurl")
 
 
cred <- OAuthFactory$new(consumerKey="XXXXXXXXXXXXXX",
                         consumerSecret="YYYYYYYYYYYYYYYYY",
                         requestURL="https://api.twitter.com/oauth/request_token",
                         accessURL="http://api.twitter.com/oauth/access_token",
                         authURL="http://api.twitter.com/oauth/authorize")
 
 
cred$handshake(cainfo=cainfo)
 
 
#save(cred, file="twitteR_credentials")
 
#load("twitteR_credentials")
 
registerTwitterOAuth(cred)
 
 
z <- getCurRateLimitInfo(cainfo=cainfo)
z$getRemainingHits()

 

이렇게 등록된 키를 이용해서 몇 가지 예제를 보여줬다. 

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73

# 팔로워 분포  -----------------------------------------------------------------
 
 
 
gogamza <- getUser("gogamza",cainfo=cainfo)
gogamza.friends <- gogamza$getFriends(cainfo=cainfo)
#gogamza.friendsID <- gogamza$getFriendIDs()
gogamza.followers <- gogamza$getFollowers(cainfo=cainfo)
 
 
fcnt <- c()
idnames <- c()
location <- c()
ids <- c()
friends <- c()
statuscnt <- c()
for(i in gogamza.followers){
  fcnt <- append(fcnt, i$followersCount)
  idnames <- append(idnames, i$name)
  location <- append(location, i$location)
  ids <- append(ids, i$id)
  friends <- append(friends, i$friendsCount)
  statuscnt <- append(statuscnt, i$statusesCount)
}
 
 
gogamza.followers.df <- data.frame(id=ids, names=idnames,followers=fcnt, friends=friends, statusCount=statuscnt)
 
cor(gogamza.followers.df[,c(3,4)])
 
library(ggplot2)
#plot graph 
fol <- ggplot(gogamza.followers.df, aes(followers, friends)) + 
  geom_point(aes(size=statusCount), alpha=I(0.8), colour="red") + 
  geom_text(aes(label=names, size=followers + friends), hjust=1, vjust=-1, angle=25) + scale_size(range=c(3,7))
 
 
# 나를 팔로잉 하는  사람들의 팔로워 분포 
 
ggplot(gogamza.followers.df) +  geom_histogram(aes(log10(followers), fill="log of followers"), alpha=I(0.7)) + 
  geom_histogram(aes(log10(friends), fill="log of friends"), alpha=I(0.7)) + scale_fill_hue("팔로워 그리고 친구")
 
#전체 그래프 
tiff(file="followers.tiff" ,width = 800, height = 600, units = "px",type="windows", family="Dotum", antialias="cleartype" ,compression="lzw")
fol
dev.off()
 
 
gogamza.followers.df.medium <- subset(gogamza.followers.df, followers <= 20000)
 
# 팔로워 2만명 이하 
tiff(file="followers2.tiff" ,width = 800, height = 600, units = "px",type="windows", family="Dotum", antialias="cleartype" ,compression="lzw")
fo2 <- fol %+% gogamza.followers.df.medium
fo2 + geom_smooth()
dev.off()
 
gogamza.followers.df.small <- subset(gogamza.followers.df, followers <= 6000)
 
# 팔로워 6천명 이하 
tiff(file="followers3.tiff" ,width = 800, height = 600, units = "px",type="windows", family="Dotum", antialias="cleartype" ,compression="lzw")
 
fo3 <- fo2 %+% gogamza.followers.df.small
fo3 + geom_smooth()
dev.off()
 
# 팔로워 천명 이하 
gogamza.followers.df.tiny <- subset(gogamza.followers.df, followers <= 1000)
 
tiff(file="followers4.tiff" ,width = 800, height = 600, units = "px",type="windows", family="Dotum", antialias="cleartype" ,compression="lzw")
 
fo4 <- fo3 %+% gogamza.followers.df.tiny
fo4 + geom_smooth()
dev.off()

 

두번째 SNA 예제 

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31

library(igraph)
gogamza <- getUser("gogamza",cainfo=cainfo)
 
gogamza.followers <- gogamza$getFollowers(cainfo=cainfo)
 
followerscnt <- sapply(gogamza.followers, function(x) {x$followersCount})
 
followers <- sapply(gogamza.followers, function(x) {x$name})
 
#팔로워가 많은 순으로 정렬해서 top 20명만 추출 
followers.df <- data.frame(followers, followerscnt)
top_followers <- order(followers.df, decreasing=T)[1:20]
 
 
followermatrix <- as.matrix(data.frame(followers=followers[top_followers], me="gogamza"))
 
# secondfollowers <- data.frame()
# 
# for(i in gogamza.followers[top_followers]){
#   foll <- i$getFollowers(cainfo=cainfo)
#   followerscnt <- sapply(foll, function(x) {x$followersCount})
#   followers <- sapply(foll, function(x) {x$name})
#   idx <- order(data.frame(followers, followerscnt), decreasing=T)[1:5]
#   secondfollowers <- rbind(secondfollowers,data.frame(sapply(foll[idx], name), i$name))
# }
 
 
sna <- graph.edgelist(followermatrix, directed=F)
 
 
plot(sna,vertex.label=V(sna)$name, layout=layout.random,  vertex.label.cex=1, vertex.label.family="Verdana")

마지막 예제는 최근 300개의 트윗을 가져와 명사를 추출한 뒤 이를 워드클라우드로 만드는 예제이다. 

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32

library(tm)
library(KoNLP)
library(wordcloud)
 
gogamza <- userTimeline(user="gogamza", n=300, cainfo=cainfo)
 
gogamzatw <- c()
for(i in 1:length(gogamza)){
  gogamzatw <- append(gogamzatw, gogamza[[i]]$text)
}
 
 
gogamzatw <- gsub("[[:space:]]"," ", gogamzatw)
 
gogamzaNoun <- sapply(gogamzatw, extractNoun,USE.NAMES=F)
 
gogamzaNoun <- unlist(gogamzaNoun, use.name=F)
 
gogamzaNoun <- gogamzaNoun[-which(gogamzaNoun %in% stopwords("english"))]
gogamzaNoun <- Filter(function(x){nchar(x)>=2}, gogamzaNoun)
gogamzaNoun <- gsub("[[:punct:]]", "", gogamzaNoun)
 
#gogamzaNoun <- Filter(function(x){nchar(x)>=2}, gogamzaNoun)
 
 
 
wordcount <- table(gogamzaNoun)
pal <- brewer.pal(8,"Dark2")
 
 
wordcloud(names(wordcount),freq=wordcount,scale=c(4,0.3),min.freq=10,
          random.order=T,rot.per=.1,colors=pal)

CC BY-NC 4.0 4회 R meetup 코드 by from __future__ import dream is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.