R Notebook

vec.name <- c("김지훈 ", "이유진 ", "박동현 ","김민지 ")
#vec.name <- c("a ", "b ", "c ","d ")
vec.name
vec.english <- c(90 , 80 , 60 , 70)
vec.english
vec.math <- c(50,60,100,20)
vec.math
vec.class <- c(1,1,2,2)
df_score <- data.frame(vec.name,vec.english,vec.math,vec.class)
df_score

id <- c(1:10)
id
class <- c(1,1,1,1,1,2,2,2,2,2)
class
math <- c(50,60,45,30,25,50,67,90,87,77)
math
english <- c(98,97,86,98,80,89,50,67,90,87)
english
science <- c(50,60,78,58,65,98,50,67,90,87)
science
df.exam <- data.frame(id,class,math ,english,science)
df.exam
#주석 :전체학생점수 보기 
df.exam
#n번째 row까지만 보여주기 head(df,n)(자주 사용됨)
head(df.exam,8)
#밑에서 n번째 row까지만 보여주기 tail(df,n)
tail(df.exam,8)
#뷰어창에서 df확인 View(df)* view의 v는 대문자이다.
View(df.exam)
#df, rowcount,col count
dim(df.exam)
#str()strucre를 나타내는 함수 (자주 사용됨) 
str(df.exam)
#summary() 요약
summary(df.exam)
#package
#install.packages("ggplot2") #전체 플젝으로 불러오느것 
#library(ggplot2)  #이 파일로 불로오는 것 
#df.mpg <- as.data.frame(ggplot2::mpg)
#df.mpgdf

df.exam %>%  #ctrl+shift+m 파이프라이
  data.table::setnames(
  )
if("dplyr" %in% installed.packages("dplyr") == FALSE)install.packages("dplyr")
library(dplyr)
df.exam %>%   #ctrl+shift+m 파이프라이
  data.table::setnames(
    old = "id",
    new="아이디 "
  )
df.exam

df.exam$total <- df.exam$math+df.exam$english+df.exam$science
df.exam 


df.exam$mean <- df.exam$total/3
df.exam 

hist(df.exam$total)
hist(df.exam$mean)
hist(df.exam$math)
#ifelse()
df.exam$test <- ifelse(df.exam$mean>=80, "합격","불합격")
df.exam

#table()
table(df.exam$mean)
library(ggplot2)
ggplot2::qplot(df.exam$test)
df.exam$class <- ifelse(df.exam$mean>=80,"A",
                    ifelse(df.exam$mean>=70,"B","C")
                    )
df.exam
ggplot2::qplot(df.exam$class)

####4함수::head ,tail, str ,summary
head(df.mpg)
tail(df.mpg)
str(df.mpg)
summary(df.mpg)
#변수명 수정 
library(dplyr)
df.mpg %>%   #ctrl+shift+m 파이프라이
  data.table::setnames(
    old = "manufacturer","model","displ","year","cyl ","trans","drv","cty","hwy","fl","class",
    new="제조회사","자동차모델","배기량","생산연도","실린더 개수","변속기 종류","구동방식"  ,"도시연비 ","고속도로 연비","연료종류","자동차 종류"
  )
df.mpg

# 파생변수 생성 
df.mpg$total <-(df.mpg$cty+df.mpg$hwy) 
df.mpg
df.mpg$mean <- (df.mpg$total)/2
df.mpg
df.mpg$test <- ifelse(df.mpg$total>=40, "pass","fail")
df.mpg

#빈도 확인
table(df.mpg$test)
library(ggplot2)
ggplot2::qplot(df.mpg$test)


#####df.midwest
library(ggplot2)
df.midwest <- as.data.frame(ggplot2::midwest)
df.midwest
head(df.midwest)
tail(df.midwest)
str(df.midwest)
summary(df.midwest)

#변수 이름 수정 하기 

df.midwest %>%    
  data.table::setnames(
    old = "poptotal","popasian",
    new="total","asian"
)
df.midwest

# 파생변수 생성 

df.midwest$aa <- (df.midwest$popasian/df.midwest$poptotal)*100
df.midwest
hist(df.midwest$aa)
mean(df.midwest$aa)
df.midwest$test <- ifelse(df.midwest$aa>0.4872462,"large","small")
df.midwest























#'data.frame':  234 obs. of  11 variables:
#  $ manufacturer(제조회사): chr  "audi" "audi" "audi" "audi" ...
#$ model  (자동차모델 )     : chr  "a4" "a4" "a4" "a4" ...
#$ displ (배기량 )      : num  1.8 1.8 2 2 2.8 2.8 3.1 1.8 1.8 2 ...
#$ year  (생산연도 )      : int  1999 1999 2008 2008 1999 1999 2008 1999 1999 2008 ...
#$ cyl    (실린더 개수 )     : int  4 4 4 4 6 6 6 4 4 4 ...
#$ trans (변속기 종류 )      : chr  "auto(l5)" "manual(m5)" "manual(m6)" "auto(av)" ...
#$ drv  (구동방식 )       : chr  "f" "f" "f" "f" ...
#$ cty  (도시연비 )       : int  18 21 20 21 16 18 18 18 16 20 ...
#$ hwy  (고속도로 연비 )       : int  29 29 31 30 26 26 27 26 25 28 ...
#$ fl     (연료종류 )     : chr  "p" "p" "p" "p" ...
#$ class  (자동차 종류 )     : chr  "compact" "compact" "compact" "compact" ...


#manufacturer          model          
#Length:234         Length:234        
#Class :character   Class :character  
#Mode  :character   Mode  :character  


#manufacturer          model          
#Length:234         Length:234        
#Class :character   Class :character  
#Mode  :character   Mode  :character  



#displ            year           cyl       
#Min.   :1.600   Min.   :1999   Min.   :4.000  
#1st Qu.:2.400   1st Qu.:1999   1st Qu.:4.000  
#Median :3.300   Median :2004   Median :6.000  
#Mean   :3.472   Mean   :2004   Mean   :5.889  
#3rd Qu.:4.600   3rd Qu.:2008   3rd Qu.:8.000  
#Max.   :7.000   Max.   :2008   Max.   :8.000  
#trans               drv           
#Length:234         Length:234        
#Class :character   Class :character  
#Mode  :character   Mode  :character  



#cty             hwy             fl           
#Min.   : 9.00   Min.   :12.00   Length:234        
#1st Qu.:14.00   1st Qu.:18.00   Class :character  
#Median :17.00   Median :24.00   Mode  :character  
#Mean   :16.86   Mean   :23.44                     
#3rd Qu.:19.00   3rd Qu.:27.00                     
#Max.   :35.00   Max.   :44.00                     
#class          
#Length:234        
#Class :character  
#Mode  :character