#05. 데이터 다루기 2 [ 실제 데이터 ]

데이터 다루기 2 [ 실제 데이터 ]

실습은 알까기2 를 바탕으로 하였습니다.

1. hflight

install.packages("hflights")

library("hflights")

hflights

str(hflights)

head(hflights)

names(hflights)

# Year Month DayofMonth DayOfWeek DepTime ArrTime UniqueCarrier FlightNum TailNum ActualElapsedTime

# 5424 2011     1          1         6    1400    1500            AA       428  N576AA                60

# 5425 2011     1          2         7    1401    1501            AA       428  N557AA                60

# 5426 2011     1          3         1    1352    1502            AA       428  N541AA                70

# 5427 2011     1          4         2    1403    1513            AA       428  N403AA                70

# 5428 2011     1          5         3    1405    1507            AA       428  N492AA                62

# 5429 2011     1          6         4    1359    1503            AA       428  N262AA                64

# AirTime ArrDelay DepDelay Origin Dest Distance TaxiIn TaxiOut Cancelled CancellationCode Diverted

# 5424      40      -10        0    IAH  DFW      224      7      13         0                         0

# 5425      45       -9        1    IAH  DFW      224      6       9         0                         0

# 5426      48       -8       -8    IAH  DFW      224      5      17         0                         0

# 5427      39        3        3    IAH  DFW      224      9      22         0                         0

# 5428      44       -3        5    IAH  DFW      224      9       9         0                         0

# 5429      45       -7       -1    IAH  DFW      224      6      13         0                         0

TableOfDest <- table(hflights$Dest)  #도착지 별 Count 

barplot(TableOfDest)

length(TableOfDest)  # 116 개의 도시가 돈재.

range(TableOfDest)  # 1~9820 Cnt 범위.

TableOfDest[TableOfDest > 6000] # 6000회 이상 도착한 곳.

TableOfDest[TableOfDest == max(TableOfDest)] # DAL 9820회.

TableOfDest[TableOfDest == min(TableOfDest)] # AGS 1회.

Over6000 <- TableOfDest[TableOfDest > 6000]

barplot(Over6000[-6])

2. Cancer 

# 나이에 따른 환자수 

DF <- read.csv("example_cancer.csv",stringsAsFactors = F)

head(DF)

# age sex height weight dateOfoperation cancerStaging hospitalization diseaseCode

# 1  75  남    161     64      2011-06-22             I              48        C187

# 2  52  여  176.6   75.3      2011-05-19            IV              17        C187

# 3  67  여    154   65.6      2011-05-31           III              10        C187

# 4  62  남    162     57      2011-06-21             I              11        C187

# 5  70  남    171     65      2011-06-09            II              10        C182

# 6  76  여    171     87      2011-06-16           III              10        C189

DegreeAge <- table(cut(DF$age,breaks = (1:11)*10))

names(DegreeAge) <- c("10s","20s","30s","40s","50s","60s","70s","80s","90s","100s")

barplot(DegreeAge)

library("ggplot2")

install.packages("ggthemes")

library("ggthemes")

De <- ggplot(data=DF , aes(x=age))

De + geom_freqpoly(binwidth=10, size=1.4, colour="orange") + theme_wsj()

3. Caffee 

install.packages("data.table")   #data.frame 과 비슷하지만 처리속도가 빨라 빅데이터에 많이 사용 된다.

library("data.table")  

library("ggplot2")

DF <- fread("example_coffee.csv", header = T,stringsAsFactors = T,data.table = F)  # read.csv 보다 빨리 데이터를 가지고온다.

head(DF)

str(DF)

class(DF)

DF <- subset(DF,select=c(-adress,-adressBystreet,-dateOfclosure,-startdateOfcessation,-duedateOfcessation,-dateOfreOpen,-zip))

names(DF)

range(DF$yearOfStart,na.rm = T)

subset(DF,subset = (yearOfStart == min(DF$yearOfStart,na.rm = T)))

unique(DF$stateOfbusiness)

#현존하는 가장 오래된 커피샵

DFFilter <- subset(DF,subset = (stateOfbusiness=="운영중"))

range(DFFilter$yearOfStart,na.rm = T)

subset(DFFilter,subset = (yearOfStart == min(DFFilter$yearOfStart,na.rm = T)))

#매년 생겨나는 카페 수.

table(DF$yearOfStart)

barplot(table(DF$yearOfStart))

qplot(yearOfStart,data=DF,geom="bar",binwidth=1)

#연도에 따른 영업상태 분할.

Freq <- table(DF$stateOfbusiness,DF$yearOfStart)

Freq <- Freq[,which(colnames(Freq) >= "1997")]

PFreq <- prop.table(Freq,margin = 2)  

?prop.table

colnames(Freq)

NewDF <- data.frame(colnames(Freq),Freq[1,],Freq[2,],PFreq[1,],PFreq[2,])

rownames(NewDF) <- NULL

colnames(NewDF) <- c("Time","Open","Close","POpen","PClose")

NewDF <- NewDF[-19,]

GP <- ggplot(NewDF, aes(x=factor(Time),y=Close,group = 1))

GP + geom_line(colour="steelblue1", size=1) + geom_point(colour="steelblue",size=3) + geom_line(aes(y=Open), colour="tomato2",size=1) + geom_point(aes(y=Open), colour = "red", size = 3) + theme_bw()

4. Size of Coffee

Size <- DF$sizeOfsite

summary(Size)

#  Min.  1st Qu.   Median     Mean  3rd Qu.     Max.     NA's 

#  0.00    28.12    50.00    75.53    93.75 24080.00       19

plot(Size) # 10000 이상의 특이한 점 발견.

#아웃라이어 ( 특이점 ) 제거 )

Size[Size > 10000] <-  NA

summary(Size)

# Min. 1st Qu.  Median    Mean 3rd Qu.    Max.    NA's 

# 0.00   28.12   50.00   75.02   93.75 1406.00      20 

Size[Size==0] <- NA

Size <- Size[complete.cases(Size)]

summary(Size)

# Min. 1st Qu.  Median    Mean 3rd Qu.    Max. 

# 0.25   30.00   51.92   77.23   95.30 1406.00 

DegreeOfSize <- table(cut(Size,breaks=c(0:72)*20))  # 20단위로 계급생성.

library("ggplot2")

library("ggthemes")

GE <- ggplot(data=DF,aes(x=sizeOfsite))

GE + geom_freqpoly(binwidth=10, size=1.2, colour="orange") + scale_x_continuous(limits = c(0,300), breaks=seq(0,300,20)) + theme_wsj()

head(DF$sizeOfsite)

5. Population Of Korea

############### 전국 인구조사 자료정리 ( 전처리 ) ################

DF <- read.csv("example_population.csv",stringsAsFactors = F)

str(DF)

head(DF,5)

# Provinces      city Population Households PersInHou   Male Female SexRatio

# 2  서울특별시   종로구      155695      72882      2.14  76962  78733     0.98

# 3  서울특별시     중구      126817      59614      2.13  63292  63525     1.00

# 4  서울특별시   용산구      235186     108138      2.17 114119 121067     0.94

# 5  서울특별시   성동구      298145     126915      2.35 148265 149880     0.99

install.packages("stringr")

library("stringr")

library("ggplot2")

#str_split_fixed(문자열, 분할 기준문자,분리할 개수)

temp <- str_split_fixed(DF[,1],"\\(",2)

NewCity <- str_split_fixed(temp[,1]," ",2)

colnames(NewCity) <- c("Provinces","city")

head(NewCity)

DF <- data.frame(NewCity,DF[,c(2:7)])

DF[DF==" "] <- NA

DF <- DF[complete.cases(DF),]

head(DF,10)

for(i in 3:8){

  DF[,i] <- sapply(DF[,i], function(x) gsub(",","",x))

  DF[,i] <- as.character(DF[,i])

  DF[,i] <- as.numeric(DF[,i])

}

str(DF)

ProPopul <- tapply(DF$Population, DF$Provinces, sum)

DF[,1] <- factor(DF[,1])

ProPopul <- tapply(DF$Population, DF$Provinces, sum)

Graph <- ggplot(DF,aes(x=Provinces,y=Population, fill=Provinces))

Graph + geom_bar(stat="identity") + theme_wsj()