2. R markdown( 변수별 EDA 및 abline 2가지 사용)
2019. 2. 1. 09:19
- EDA(Explanantory Data analysis)
데이터 준비 및 확인
library(moonBook)
data(acs) # 급성 심근경색
#View(acs)
dim(acs)## [1] 857 17head(acs)## age sex cardiogenicShock entry Dx EF height weight
## 1 62 Male No Femoral STEMI 18.0 168 72
## 2 78 Female No Femoral STEMI 18.4 148 48
## 3 76 Female Yes Femoral STEMI 20.0 NA NA
## 4 89 Female No Femoral STEMI 21.8 165 50
## 5 56 Male No Radial NSTEMI 21.8 162 64
## 6 73 Female No Radial Unstable Angina 22.0 153 59
## BMI obesity TC LDLC HDLC TG DM HBP smoking
## 1 25.51020 Yes 215 154 35 155 Yes No Smoker
## 2 21.91381 No NA NA NA 166 No Yes Never
## 3 NA No NA NA NA NA No Yes Never
## 4 18.36547 No 121 73 20 89 No No Never
## 5 24.38653 No 195 151 36 63 Yes Yes Smoker
## 6 25.20398 Yes 184 112 38 137 Yes Yes Neverstr(acs)## 'data.frame': 857 obs. of 17 variables:
## $ age : int 62 78 76 89 56 73 58 62 59 71 ...
## $ sex : chr "Male" "Female" "Female" "Female" ...
## $ cardiogenicShock: chr "No" "No" "Yes" "No" ...
## $ entry : chr "Femoral" "Femoral" "Femoral" "Femoral" ...
## $ Dx : chr "STEMI" "STEMI" "STEMI" "STEMI" ...
## $ EF : num 18 18.4 20 21.8 21.8 22 24.7 26.6 28.5 31.1 ...
## $ height : num 168 148 NA 165 162 153 167 160 152 168 ...
## $ weight : num 72 48 NA 50 64 59 78 50 67 60 ...
## $ BMI : num 25.5 21.9 NA 18.4 24.4 ...
## $ obesity : chr "Yes" "No" "No" "No" ...
## $ TC : num 215 NA NA 121 195 184 161 136 239 169 ...
## $ LDLC : int 154 NA NA 73 151 112 91 88 161 88 ...
## $ HDLC : int 35 NA NA 20 36 38 34 33 34 54 ...
## $ TG : int 155 166 NA 89 63 137 196 30 118 141 ...
## $ DM : chr "Yes" "No" "No" "No" ...
## $ HBP : chr "No" "Yes" "Yes" "No" ...
## $ smoking : chr "Smoker" "Never" "Never" "Never" ...summary(acs)## age sex cardiogenicShock entry
## Min. :28.00 Length:857 Length:857 Length:857
## 1st Qu.:55.00 Class :character Class :character Class :character
## Median :64.00 Mode :character Mode :character Mode :character
## Mean :63.31
## 3rd Qu.:72.00
## Max. :91.00
##
## Dx EF height weight
## Length:857 Min. :18.00 Min. :130.0 Min. : 30.00
## Class :character 1st Qu.:50.45 1st Qu.:158.0 1st Qu.: 58.00
## Mode :character Median :58.10 Median :165.0 Median : 65.00
## Mean :55.83 Mean :163.2 Mean : 64.84
## 3rd Qu.:62.35 3rd Qu.:170.0 3rd Qu.: 72.00
## Max. :79.00 Max. :185.0 Max. :112.00
## NA's :134 NA's :93 NA's :91
## BMI obesity TC LDLC
## Min. :15.62 Length:857 Min. : 25.0 Min. : 15.0
## 1st Qu.:22.13 Class :character 1st Qu.:154.0 1st Qu.: 88.0
## Median :24.16 Mode :character Median :183.0 Median :114.0
## Mean :24.28 Mean :185.2 Mean :116.6
## 3rd Qu.:26.17 3rd Qu.:213.0 3rd Qu.:141.0
## Max. :41.42 Max. :493.0 Max. :366.0
## NA's :93 NA's :23 NA's :24
## HDLC TG DM HBP
## Min. : 4.00 Min. : 11.0 Length:857 Length:857
## 1st Qu.:32.00 1st Qu.: 68.0 Class :character Class :character
## Median :38.00 Median :105.5 Mode :character Mode :character
## Mean :38.24 Mean :125.2
## 3rd Qu.:45.00 3rd Qu.:154.0
## Max. :89.00 Max. :877.0
## NA's :23 NA's :15
## smoking
## Length:857
## Class :character
## Mode :character
##
##
##
## apply(is.na(acs), MARGIN = 2, FUN='sum')## age sex cardiogenicShock entry
## 0 0 0 0
## Dx EF height weight
## 0 134 93 91
## BMI obesity TC LDLC
## 93 0 23 24
## HDLC TG DM HBP
## 23 15 0 0
## smoking
## 0library(mlr)## Loading required package: ParamHelperssummarizeColumns(acs)## name type na mean disp median
## 1 age integer 0 63.31155 11.69630382 64.00000
## 2 sex character 0 NA 0.33488915 NA
## 3 cardiogenicShock character 0 NA 0.06067678 NA
## 4 entry character 0 NA 0.36406068 NA
## 5 Dx character 0 NA 0.53325554 NA
## 6 EF numeric 134 55.83444 9.62296368 58.10000
## 7 height numeric 93 163.17539 9.07930672 165.00000
## 8 weight numeric 91 64.84373 11.35519114 65.00000
## 9 BMI numeric 93 24.28190 3.34669608 24.16104
## 10 obesity character 0 NA 0.33838973 NA
## 11 TC numeric 23 185.20024 47.77292477 183.00000
## 12 LDLC integer 24 116.58343 41.09484883 114.00000
## 13 HDLC integer 23 38.23501 11.08668982 38.00000
## 14 TG integer 15 125.23872 90.85259401 105.50000
## 15 DM character 0 NA 0.35472579 NA
## 16 HBP character 0 NA 0.41540257 NA
## 17 smoking character 0 NA 0.61260210 NA
## mad min max nlevs
## 1 13.343400 28.0000 91.00000 0
## 2 NA 287.0000 570.00000 2
## 3 NA 52.0000 805.00000 2
## 4 NA 312.0000 545.00000 2
## 5 NA 153.0000 400.00000 3
## 6 7.857780 18.0000 79.00000 0
## 7 7.413000 130.0000 185.00000 0
## 8 10.378200 30.0000 112.00000 0
## 9 3.011608 15.6157 41.42012 0
## 10 NA 290.0000 567.00000 2
## 11 42.995400 25.0000 493.00000 0
## 12 40.030200 15.0000 366.00000 0
## 13 10.378200 4.0000 89.00000 0
## 14 60.045300 11.0000 877.00000 0
## 15 NA 304.0000 553.00000 2
## 16 NA 356.0000 501.00000 2
## 17 NA 204.0000 332.00000 3VIM 패키지의 VIM 라이브러리를 통한 null값 확인
# install.packages("VIM")
library(VIM)## Loading required package: colorspace## Loading required package: grid## Loading required package: data.table## VIM is ready to use.
## Since version 4.0.0 the GUI is in its own package VIMGUI.
##
## Please use the package to use the new (and old) GUI.## Suggestions and bug-reports can be submitted at: https://github.com/alexkowa/VIM/issues##
## Attaching package: 'VIM'## The following object is masked from 'package:datasets':
##
## sleepaggr(acs, prop = F, numbers = T)
범주형/숫자형 변수 나누어서 분석해주는 라이브러리 xda
# 범주형/숫자형 변수 확인을 위한 패키지 xda 설치
#https://github.com/ujjwalkarn/xda
# library(devtools)
# install_github("ujjwalkarn/xda")
# 설치방법 : 1) *관리자 권환으로실행 -> 2) 깃허브인스톨 패키지설치 -> 3) 깃허브인스톨라이브러리로 xda 를 다운 -> 4) xda 라이브러리 사용
# install.packages("githubinstall")
# library(githubinstall)
# githubinstall("xda")
# 사용
# library(xda)
# 데이터셋 자체에 대해서 범주형과 숫자형 변수(칼럼)들 자동분석
# bivariate(iris,'Species','Sepal.Length',n.bins=2)
# bivariate(iris,'Sepal.Length','Species')범주형 변수 분석 (범주 -> 빈도 / 범주-범주 -> 빈도로 이루어진 교차표 )
- table()함수로 범주형 1개(단변수), 범주형 2개(이변수)를 빈도/교차표 생성
- 카이제곱검정 chisq.test( table = 교차표 )로 범주형변수 2개의 연관성 확인
- H0 : 서로 독립이다.
# 1. 범주형 단변수 분석
table(acs$Dx) # 범주별 빈도 확인##
## NSTEMI STEMI Unstable Angina
## 153 304 400prop.table( table(acs$Dx)) # 범주별 상대빈도 확인##
## NSTEMI STEMI Unstable Angina
## 0.1785298 0.3547258 0.4667445barplot( table(acs$Dx) ,
col = c("orange", "blue", "red") ) # 범주별 빈도 bar chart
# install.packages("qualityTools")
library(qualityTools)## Loading required package: Rsolnp##
## Attaching package: 'Rsolnp'## The following object is masked from 'package:mlr':
##
## benchmark## Loading required package: MASS##
## Attaching package: 'qualityTools'## The following object is masked from 'package:data.table':
##
## cube## The following object is masked from 'package:stats':
##
## sigmaparetoChart(table(acs$Dx)) # 범주별빈도 + 누적합계 bar chart + line 
##
## Frequency 400 304 153
## Cum. Frequency 400 704 857
## Percentage 46.7% 35.5% 17.9%
## Cum. Percentage 46.7% 82.1% 100.0%##
## Frequency 400.00000 304.00000 153.00000
## Cum. Frequency 400.00000 704.00000 857.00000
## Percentage 46.67445 35.47258 17.85298
## Cum. Percentage 46.67445 82.14702 100.00000paretoChart(table(acs$Dx),
showTable = F)
##
## Frequency 400 304 153
## Cum. Frequency 400 704 857
## Percentage 46.7% 35.5% 17.9%
## Cum. Percentage 46.7% 82.1% 100.0%##
## Frequency 400.00000 304.00000 153.00000
## Cum. Frequency 400.00000 704.00000 857.00000
## Percentage 46.67445 35.47258 17.85298
## Cum. Percentage 46.67445 82.14702 100.00000# 2. 범주형 이변수 분석
table(acs$sex, acs$Dx) # 범주-범주 분석- 범주별 빈도 교차표##
## NSTEMI STEMI Unstable Angina
## Female 50 84 153
## Male 103 220 247tab <-table(acs$sex, acs$Dx)
addmargins(tab) # 범주2개의 교차표에 sum을 표함시킨다.##
## NSTEMI STEMI Unstable Angina Sum
## Female 50 84 153 287
## Male 103 220 247 570
## Sum 153 304 400 857prop.table(tab) # 전체합에 대한 상대빈도 교차표##
## NSTEMI STEMI Unstable Angina
## Female 0.05834306 0.09801634 0.17852975
## Male 0.12018670 0.25670945 0.28821470prop.table(tab, 1) # 1= 가로에 있는 범주(sex)에 대한(분모) 상대빈도 교차표##
## NSTEMI STEMI Unstable Angina
## Female 0.1742160 0.2926829 0.5331010
## Male 0.1807018 0.3859649 0.4333333prop.table(tab, 2) # 2=0=세로에 있는 범주(Dx)에 대한 상대빈도 교차표##
## NSTEMI STEMI Unstable Angina
## Female 0.3267974 0.2763158 0.3825000
## Male 0.6732026 0.7236842 0.6175000chisq.test( table(acs$sex, acs$Dx)) # 2개 범주의 독립성 확인(H0 : 독립)##
## Pearson's Chi-squared test
##
## data: table(acs$sex, acs$Dx)
## X-squared = 8.7983, df = 2, p-value = 0.01229mycol <- c("skyblue", "blue") # 범주 2개에 대한 색 미리 지정
barplot( table(acs$sex, acs$Dx), # 앞에 넣은 범주가 범례가 된다.
col = mycol , main ="Segmented Bar Plot of Gender")
legend("topleft", c("females "," males "), col= mycol, pch = 16)
mycol<-c("skyblue", "blue","darkblue")
barplot(table(acs$Dx, acs$sex), col = mycol, main = "Segmented Bar Plot of Diagnosis")
legend("topleft", c("NSTEMI", "STEMI", "Unstable Angina"), col= mycol, pch = 16, inset =0.05)
숫자형 변수 분석
- 숫자형 변수는 분포를 확인해야하니 boxplot 이다.
- 숫자형 변수의 분포 확인2 - 정규분포확인 hist
# 1. 숫자형 단변수
summary(acs$TG)## Min. 1st Qu. Median Mean 3rd Qu. Max. NA's
## 11.0 68.0 105.5 125.2 154.0 877.0 15quantile(acs$TG, na.rm=T)## 0% 25% 50% 75% 100%
## 11.0 68.0 105.5 154.0 877.0IQR(acs$TG, na.rm=T)## [1] 86var(acs$TG, na.rm=T)## [1] 8254.194sd(acs$TG, na.rm=T)## [1] 90.85259stem(acs$TG) # 줄기-잎 플롯##
## The decimal point is 2 digit(s) to the right of the |
##
## 0 | 12222222233333333333333333333333333333444444444444444444444444444444
## 0 | 55555555555555555555555555555555555555555555556666666666666666666666+220
## 1 | 00000000000000000000000000000000000000000000000111111111111111111111+151
## 1 | 55555555555555555555555555555555555666666666666666666666666666666777+55
## 2 | 000000001111111112222222222222222333344444444444
## 2 | 555556666666667778888899
## 3 | 000111223
## 3 | 5788899
## 4 | 1123
## 4 | 555669
## 5 | 4
## 5 | 58
## 6 | 2
## 6 | 58
## 7 |
## 7 |
## 8 | 0
## 8 | 8#install.packages("mosaic")
library(mosaic)## Loading required package: dplyr##
## Attaching package: 'dplyr'## The following object is masked from 'package:MASS':
##
## select## The following objects are masked from 'package:data.table':
##
## between, first, last## The following objects are masked from 'package:stats':
##
## filter, lag## The following objects are masked from 'package:base':
##
## intersect, setdiff, setequal, union## Loading required package: lattice##
## Attaching package: 'lattice'## The following object is masked from 'package:moonBook':
##
## densityplot## Loading required package: ggformula## Loading required package: ggplot2## Loading required package: ggstance##
## Attaching package: 'ggstance'## The following objects are masked from 'package:ggplot2':
##
## geom_errorbarh, GeomErrorbarh##
## New to ggformula? Try the tutorials:
## learnr::run_tutorial("introduction", package = "ggformula")
## learnr::run_tutorial("refining", package = "ggformula")## Loading required package: mosaicData## Loading required package: Matrix##
## The 'mosaic' package masks several functions from core packages in order to add
## additional features. The original behavior of these functions should not be affected by this.
##
## Note: If you use the Matrix package, be sure to load it BEFORE loading mosaic.##
## Attaching package: 'mosaic'## The following object is masked from 'package:Matrix':
##
## mean## The following object is masked from 'package:ggplot2':
##
## stat## The following objects are masked from 'package:dplyr':
##
## count, do, tally## The following object is masked from 'package:qualityTools':
##
## dotPlot## The following object is masked from 'package:mlr':
##
## resample## The following objects are masked from 'package:stats':
##
## binom.test, cor, cor.test, cov, fivenum, IQR, median,
## prop.test, quantile, sd, t.test, var## The following objects are masked from 'package:base':
##
## max, mean, min, prod, range, sample, sumdotPlot(acs$TG,pch=19,col=6) # Dot plot
hist(acs$TG, pch=19, col="lightgray") # 히스토 그램 with 평균, 중앙값
abline(v= mean(acs$TG,na.rm=T), col = " blue")
abline(v= median (acs$TG,na.rm=T), col = "green")
legend("topright", c("Mean", "Median"), pch = c(16,15), col = c("blue", "green"))
par(mfrow=c(1,2))
boxplot(acs$TG, notch = T, col="orange", main="TG (in original scale)")
boxplot(log(acs$TG), notch = T, col="orange", main="TG (in log scale)")
par(mfrow=c(1,1))
# 2. 숫자형 이변수
# 데이터 준비
data("iris")
attach(iris)
# 데이터 확인
library(mlr)
summarizeColumns(iris) # 4가지 숫자형 + 1가지 범주형## name type na mean disp median mad min max
## 1 Sepal.Length numeric 0 5.843333 0.8280661 5.80 1.03782 4.3 7.9
## 2 Sepal.Width numeric 0 3.057333 0.4358663 3.00 0.44478 2.0 4.4
## 3 Petal.Length numeric 0 3.758000 1.7652982 4.35 1.85325 1.0 6.9
## 4 Petal.Width numeric 0 1.199333 0.7622377 1.30 1.03782 0.1 2.5
## 5 Species factor 0 NA 0.6666667 NA NA 50.0 50.0
## nlevs
## 1 0
## 2 0
## 3 0
## 4 0
## 5 3boxplot(Sepal.Length, notch = TRUE, col = "grey") # 숫자형 단변수의 분포 boxplot
boxplot(Sepal.Length~Species, data=iris, notch = TRUE, col = "orange") # 범주별 숫자의 분포boxplot 
# 숫자형 변수의 분포 확인2 - hist로 정규분폰지 보기
hist(Sepal.Length,
col = "grey",
main = "R default",
ylab = "Frequency",
freq = FALSE)
plot(iris[,-5]) # 산점도 행렬 기본
plot(iris,'Petal.Length') # 특정 숫자칼럼에 대한 산점도 행렬 
pairs(iris[,-5], gap = 0) # 산점도행렬
pairs(iris[,-5], gap = 0, panel = panel.smooth) #산점도행렬 + 회귀라인
# 범주별 - 산점도(숫자-숫자) 행렬
# install.packages("mycor")
library(mycor)
plot(mycor(iris),type=1,groups=Species)
plot(mycor(iris),type=2,groups=Species) #hist추가
plot(mycor(iris),type=3,groups=Species) #상관계수와 fitting curve
abline의 2가지 이용기
- 숫자-숫자형의 산점도에서 나오는 회귀식 결과 or 절편(a)과 기울기(b)로 회귀선을 그림
- 수평(h) 혹은 수직(v)선을 추가로 그려줌
#abline-회귀식에서 나온 계수로 회귀선 그리기
plot (cars , xlim =c(0, 25) )
abline (a=-5, b=3.5 , col ="red ")
abline(lm(dist~speed,data=cars),col="green")
# 추가 수평/수직선 그리기
plot (cars , xlim =c(0, 25) )
abline (a=-5, b=3.5 , col ="red ")
abline (h= mean ( cars $ dist ), lty =2, col =" blue ")
abline (v= mean ( cars $ speed ), lty =2, col =" green ")
'한의대 생활 > └ 통계에 대한 나의 정리' 카테고리의 다른 글
| 3. 통계적 추론과 가설검정, p-value (0) | 2019.02.05 |
|---|---|
| 2-2 R markdown ggplot2 ( plotly 올릴시 에러 ) (0) | 2019.02.01 |
| 2. 전처리시 체크2가지 및 EDA시 변수의 성격에 따른 분류 (0) | 2019.01.25 |
| 1. R markdown(데이터경로, 불러오기, 5가지확인, summarizeColumns, mytable, mycsv) (0) | 2019.01.25 |
| 1. 통계 - 기술통계와 추론통계 , 표본추출방법들 (0) | 2019.01.19 |