Working Data Loading

load("./crimtab.RData")
ls()
## [1] "crimtab_2"       "crimtab_df"      "crimtab_long"    "crimtab_long_df"
ls.str()
## crimtab_2 :  'table' int [1:42, 1:22] 0 0 0 0 0 0 1 0 0 0 ...
## crimtab_df : 'data.frame':   924 obs. of  3 variables:
##  $ finger: num  9.4 9.5 9.6 9.7 9.8 9.9 10 10.1 10.2 10.3 ...
##  $ height: num  56 56 56 56 56 56 56 56 56 56 ...
##  $ Freq  : int  0 0 0 0 0 0 1 0 0 0 ...
## crimtab_long :  num [1:3000, 1:2] 10 10.3 9.9 10.2 10.2 10.3 10.4 10.7 10 10.1 ...
## crimtab_long_df : 'data.frame':  3000 obs. of  2 variables:
##  $ finger: num  10 10.3 9.9 10.2 10.2 10.3 10.4 10.7 10 10.1 ...
##  $ height: num  56 57 58 58 58 58 58 58 59 59 ...
head(crimtab_long_df)
##   finger height
## 1   10.0     56
## 2   10.3     57
## 3    9.9     58
## 4   10.2     58
## 5   10.2     58
## 6   10.3     58

Graphic Representation

Base Graphics

  • 키와 손가락길이의 산점도
# plot(finger ~ height, data = crimtab_long_df)
plot(crimtab_long_df[2:1])

  • 변수 각각의 히스토그램은?
par(mfrow = c(1, 2))
hist(crimtab_long_df$height, 
     main = "Histogram of Height", 
     xlab = "height(inches)")
hist(crimtab_long_df$finger, 
     main = "Histogram of Finger Length", 
     xlab = "finger length(cm)")

# hist(crimtab_long_df["height"], main="Histogram of Height", xlab="height(inches)")
# hist(crimtab_long_df["finger"], main="Histogram of Finger Length", xlab= "finger length(cm)")
  • 평균과 표준편차를 한번에 구하려면 다음과 같이 anonymous function을 작성하고 mapply() 또는 sapply()를 이용하는 게 편함. 이를 모수로 하는 정규곡선을 덧씌워 볼 것.
    • mean_sd()도 anonymous function 으로 평균과 표준편차를 계산해서 출력하는 함수임. 이와 같은 함수를 저장해 놓으려면 dump()를 이용함.
    • 이와 같이 계산한 평균과 표준편차를 모수로 하는 정규곡선을 덧씌워 볼 것.
mean_sd <- function(x) {
  mean <- mean(x, na.rm = TRUE)
  sd <- sd(x)
  return(c(mean = mean, sd = sd))
# list(mean = mean, sd = sd)
}
dump("mean_sd", file = "mean_sd.R")
crimtab_stat <- sapply(crimtab_long_df, mean_sd)
# crimtab_stat <- mapply(mean_sd, crimtab_long_df)
# apply(crimtab.long, 2, mean)
# apply(crimtab.long, 2, sd)
str(crimtab_stat)
##  num [1:2, 1:2] 11.547 0.549 65.473 2.558
##  - attr(*, "dimnames")=List of 2
##   ..$ : chr [1:2] "mean" "sd"
##   ..$ : chr [1:2] "finger" "height"
  • crimtab_stat이 어떤 성격을 갖는지 다음 질문과 추출 작업을 통해서 알아보자.
is.matrix(crimtab_stat)
## [1] TRUE
is.table(crimtab_stat)
## [1] FALSE
is.list(crimtab_stat)
## [1] FALSE
is.data.frame(crimtab_stat)
## [1] FALSE
crimtab_stat[, 1]
##       mean         sd 
## 11.5473667  0.5487137
crimtab_stat[, "finger"]
##       mean         sd 
## 11.5473667  0.5487137
crimtab_stat[, "finger"][1]
##     mean 
## 11.54737
crimtab_stat[, "finger"][[1]]
## [1] 11.54737
crimtab_stat[1]
## [1] 11.54737
crimtab_stat[2:3]
## [1]  0.5487137 65.4730000
# crimtab_stat["finger"]
# crimtab_stat$finger

matrix 를 data frame 으로 변환하면

(crimtab_stat_df <- data.frame(crimtab_stat))
##          finger    height
## mean 11.5473667 65.473000
## sd    0.5487137  2.557757
is.matrix(crimtab_stat_df)
## [1] FALSE
is.table(crimtab_stat_df)
## [1] FALSE
is.list(crimtab_stat_df)
## [1] TRUE
is.data.frame(crimtab_stat_df)
## [1] TRUE
crimtab_stat_df[, 1]
## [1] 11.5473667  0.5487137
str(crimtab_stat_df[, 1])
##  num [1:2] 11.547 0.549
crimtab_stat_df[, "finger"]
## [1] 11.5473667  0.5487137
str(crimtab_stat_df[, "finger"])
##  num [1:2] 11.547 0.549
crimtab_stat_df[, "finger"][1]
## [1] 11.54737
str(crimtab_stat_df[, "finger"][1])
##  num 11.5
crimtab_stat_df[, "finger"][[1]]
## [1] 11.54737
str(crimtab_stat_df[, "finger"][[1]])
##  num 11.5
crimtab_stat_df[1]
##          finger
## mean 11.5473667
## sd    0.5487137
str(crimtab_stat_df[1])
## 'data.frame':    2 obs. of  1 variable:
##  $ finger: num  11.547 0.549
crimtab_stat_df["finger"]
##          finger
## mean 11.5473667
## sd    0.5487137
str(crimtab_stat_df["finger"])
## 'data.frame':    2 obs. of  1 variable:
##  $ finger: num  11.547 0.549
crimtab_stat_df["finger"][1]
##          finger
## mean 11.5473667
## sd    0.5487137
str(crimtab_stat_df["finger"][1])
## 'data.frame':    2 obs. of  1 variable:
##  $ finger: num  11.547 0.549
crimtab_stat_df["finger"][[1]]
## [1] 11.5473667  0.5487137
str(crimtab_stat_df["finger"][[1]])
##  num [1:2] 11.547 0.549
crimtab_stat_df$finger
## [1] 11.5473667  0.5487137
str(crimtab_stat_df$finger)
##  num [1:2] 11.547 0.549
crimtab_stat_df$finger[1]
## [1] 11.54737
str(crimtab_stat_df$finger[1])
##  num 11.5
crimtab_stat_df$finger[[1]]
## [1] 11.54737
str(crimtab_stat_df$finger[[1]])
##  num 11.5

ggplot

  • 키와 손가락 길이의 산점도
library(ggplot2)
## Registered S3 methods overwritten by 'ggplot2':
##   method         from 
##   [.quosures     rlang
##   c.quosures     rlang
##   print.quosures rlang
g1 <- ggplot(data = crimtab_long_df, 
             mapping = aes(x = height, y = finger)) 
g2 <- g1 + 
  geom_point()
g2

  • 투명도 변경 : alpha = 0.9
g2.2 <- g1 + 
  geom_point(alpha = 0.9)
g2.2

  • 투명도 변경 : alpha = 0.5
g2.3 <- g1 + 
  geom_point(alpha = 0.5)
g2.3

  • 투명도 변경 : alpha = 0.1
g2.4 <- g1 + 
  geom_point(alpha = 0.1)
g2.4

  • 중복점 흐트러놓기 : position = jitter
g2.5 <- g1 + 
  geom_point(position = "jitter")
g2.5

  • 점의 크기를 줄이고 중복점 흐트러놓기 : position = jitter, size = 0.7
g2.6 <- g1 + 
  geom_point(position = "jitter", size = 0.7)
g2.6

  • 동일한 효과 : position = position_jitter(), size = 0.7
g2.7 <- g1 + 
  geom_point(position = position_jitter(), size = 0.7)
g2.7

  • 흐트러놓는 폭 조절 : width = 1, height = 0, size = 0.7
g2.8 <- g1 + 
  geom_point(position = position_jitter(width = 1, height = 0), size = 0.7)
g2.8

  • 흐트러놓는 폭과 높이 조절 : width = 1, height = 0.1, size = 0.7
g2.9 <- g1 + 
  geom_point(position = position_jitter(width = 1, height = 0.1), 
             size = 0.7)
g2.9

  • 흑백 테마 : theme_bw()
g3 <- g2.9 +
  theme_bw()
g3

히스토그램

h1 <- ggplot(data = crimtab_long_df, 
             mapping = aes(x = height)) 
h1 + geom_histogram(alpha = 0.5)
## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.

f1 <- ggplot(data = crimtab_long_df, 
             mapping = aes(x = finger))
f1 + geom_histogram(alpha = 0.5)
## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.

h1 + geom_histogram(aes(y = ..density..), 
                    binwidth = 1, 
                    alpha = 0.5)

f1 + geom_histogram(aes(y = ..density..), 
                    binwidth = 0.1, 
                    alpha = 0.5)

(g.h.1 <- h1 + 
   geom_histogram(aes(y = ..density..), 
                  binwidth = 1, 
                  fill = "white", 
                  colour = "black"))

(g.h <- g.h.1 +
   theme_bw() +
   scale_x_continuous(name = "", 
                      breaks = NULL) +
   scale_y_continuous(name = "", 
                      breaks = NULL))

(g.f.1 <- f1 + 
   geom_histogram(aes(y = ..density..), 
                  binwidth = 0.2, 
                  fill = "white", 
                  colour = "black"))

(g.f <- g.f.1 +
   theme_bw() +
   scale_x_continuous(name = "", 
                      breaks = NULL) +
   scale_y_continuous(name = "", 
                      breaks = NULL) +
   coord_flip())

평균 위치를 화살표로 나타내려면

library(grid)
(mean_finger <- crimtab_stat[, 1][[1]])
## [1] 11.54737
(sd_finger <- crimtab_stat[, 1][[2]])
## [1] 0.5487137
(mean_height <- crimtab_stat[, 2][[1]])
## [1] 65.473
(sd_height <- crimtab_stat[, 2][[2]])
## [1] 2.557757
x_finger <- seq(9.5, 13.5, 
                length.out = 3000)
y_finger <- dnorm(x_finger, 
                  mean = mean_finger, 
                  sd = sd_finger)
x_height <- seq(56, 77, 
                length.out = 3000)
y_height <- dnorm(x_height, 
                  mean = mean_height, 
                  sd = sd_height)

(g.h.2 <- g.h.1 + 
    annotate("segment", 
             x = mean_height, 
             xend = mean_height, 
             y = -0.02, 
             yend = 0, 
             arrow = arrow(length = unit(0.3, "cm"))))

(g.h.3 <- g.h.2 + 
   theme_bw())

(g.h.4 <- g.h.3 + 
   geom_line(aes(x = x_height, y = y_height)))

(g.f.2 <- g.f.1 + 
   annotate("segment", 
            x = mean_finger, 
            xend = mean_finger, 
            y = -0.1, 
            yend = 0, 
            arrow = arrow(length = unit(0.3, "cm"))))

(g.f.3 <- g.f.2 + 
   theme_bw())

(g.f.4 <- g.f.3 + 
   geom_line(aes(x = x_finger, y = y_finger)))

산점도와 히스토그램 함께 배열하기

grid 및 gridExtra 패키지와 함께 blank Grob 설정이 핵심. grid.arrange 사용법에 유의.

library(gridExtra)
blank <- grid.rect(gp = gpar(col = "white"), draw = FALSE)
grid.arrange(g.h, blank, g3, g.f, 
             ncol = 2, 
             widths = c(4, 1), 
             heights = c(1, 4))