R 기초 2

R 기초 2

파생변수 생성

cars
cars <- cars

변수이름 확인

colnames(cars)
colnames(cars) <- c("속도", "거리")
head(cars)
tail(cars)

파생변수 생성

cars$속도X거리 <- cars$속도 * cars$거리
head(cars)

조건에 따라 변수의 값을 새로 생성

cars$속도정도 <- ifelse(cars$속도 > 7, "빠름", "느림")
head(cars)

hflights 패키지 설치

install.packages('hflights')
require(hflights)
hflights <- hflights
summary(hflights)

Select

require(dplyr)
str(hflights)
newhflights <- select(hflights, Origin, Distance, TailNum)

Filter

head(filter(hflights, Month==2))
hflights1 <- filter(hflights, Month==2)
nrow(hflights1)

head(filter(hflights, Month==8))
hflights2 <- filter(hflights, Month==8)
nrow(hflights2)

hflights3 <- filter(hflights, Month==2 | Month==8)
nrow(hflights3)

hflights4 <- filter(hflights, Month==2, Dest=="DFW")
nrow(hflights4)

정렬 Arrange

hflights5 <- arrange(hflights, Month)
head(hflights5)
tail(hflights5)

hflights6 <- arrange(hflights, desc(Month))
head(hflights6)

hflights7 <- arrange(hflights, desc(Month), DayOfWeek)
head(hflights7)

열의 조작

head(
  mutate(hflights, 
         gain = ArrDelay - DepDelay,
         gain_per_hour = gain / (AirTime/60)
  )
)

Summary

summarise(hflights,
          delay = mean(DepDelay,
                       na.rm = TRUE)
)

그룹별 요약 %>%

hflights %>%
  group_by(TailNum) %>%
  select(Month, DayofMonth, DayOfWeek) %>%
  summarise(
    AveMonth = mean(Month, na.rm = TRUE),
    AveDayofMonth = mean(DayofMonth, na.rm = TRUE),
    AveDayofWeek = mean(DayOfWeek, na.rm = TRUE)
  )

연습문제 1

1

hflights8 <- select(hflights, FlightNum, TailNum, AirTime, Distance, Origin)
head(hflights8)

2

hflights9 <- filter(hflights8, Origin == "HOU", Distance >1000)
head(hflights9)

3

hflights9 %>%
  group_by(TailNum) %>%
  select(AirTime, Distance) %>%
  summarise(
    AveAirTime = mean(AirTime, na.rm = TRUE),
    AveDistance = mean(Distance, na.rm = TRUE)
  )

ggplot2

패키지 설치

install.packages('ggplot2')
require(ggplot2)

데이터 불러오기

data("midwest", package = "ggplot2")
summary(midwest)

options(scipen = 999)
ggplot(midwest, aes(x=area, y=poptotal))

Scatter Plot 생성

midwest %>%
  ggplot(aes(x=area, y=poptotal)) +
  geom_point()

리그레이션 라인 추가

midwest %>%
  ggplot(aes(x=area, y=poptotal)) +
  geom_point() +
  geom_smooth(method = "lm")

그래프의 범위 조정

g <-midwest %>%
      ggplot(aes(x=area, y=poptotal)) +
      geom_point() +
      geom_smooth(method = "lm")
g + xlim(c(0,0.1)) + ylim(c(0,1000000))

줌

g1 <- g + coord_cartesian(xlim = c(0,0.75), ylim = c(0, 500000))
g1

g1 <- g + xlim(c(0,0.1)) + ylim(c(0, 1000000))

제목과 축 이름

g1
g1 + labs(title = "Area VS Population",
          subtitle = "midwest",
          y = "Population",
          x = "Area",
          caption = "Midwest Population")

색변경

g <-midwest %>%
  ggplot(aes(x=area, y=poptotal)) +
  geom_point(col = "red", size = 1) +
  geom_smooth(method = "lm")
g + xlim(c(0,0.1)) + ylim(c(0,1000000))

색에 데이터 정보 반영

색변경

g <-midwest %>%
  ggplot(aes(x=area, y=poptotal)) +
  geom_point(aes(col=state), size = 1) +
  geom_smooth(method = "lm")
g + xlim(c(0,0.1)) + ylim(c(0,1000000))

그래프 분할

g <-midwest %>%
  ggplot(aes(x=area, y=poptotal)) +
  geom_point() +
  geom_smooth(method = "lm")
g + facet_wrap(~state, nrow = 2)

midwest
hist(midwest$popdensity)
midwest$degreeofdensity <- ifelse(midwest$popdensity >= 10000, "high", "low")
table(midwest$degreeofdensity)

g <-midwest %>%
  ggplot(aes(x=area, y=poptotal)) +
  geom_point(aes(col = degreeofdensity)) +
  geom_smooth(method = "lm")

g + facet_wrap(~state, nrow=2)

ggplot 실습

mtcars <- mtcars
summary(mtcars)
plot(mtcars)

연습문제 2

mtcars 데이터를 활용해서 두개의 연속형 변수를 선택한 후 scatter plot 그려보기

mtcars %>%
  ggplot(aes(x=mpg, y=hp)) +
  geom_point()

점에 카테고리컬 변수의 정보를 활용해서 점의 색깔에 반영하기

g <-mtcars %>%
  ggplot(aes(x=mpg, y=hp)) +
  geom_point(aes(col = as.factor(cyl)))
g

mtcars$names <- rownames(mtcars)
colnames(mtcars)
g <-mtcars %>%
  ggplot(aes(x=wt, y=mpg)) +
  geom_point(aes(col = as.factor(vs))) +
  geom_smooth(method = "lm") +
  geom_text(aes(label = names), size=3)
g