t="C:\\Users\\pc\\Downloads\\CAC NGHIEN CUU\\BAI GIANG CAC MON\\GS TUAN\\BG 12.6.19\\Diabetes data.csv"
db=read.csv(t)
head(db)

##   id age gender height weight waist hip sysbp diabp active hypertension
## 1  1  76 Female    163     53    90  93   160    90      0            1
## 2  1  40 Female    149     51    74  94   100    60      0            0
## 3  1  51 Female    151     55    91 100   120    80      0            0
## 4  1  43 Female    158     62    78  96   120    80      1            0
## 5  2  72 Female    148     47    91  95   130    60      1            0
## 6  2  44   Male    155     48    69  86   120    80      0            0
##     bmi  whr diabetes
## 1 19.95 0.97      IFG
## 2 22.97 0.79   Normal
## 3 24.12 0.91   Normal
## 4 24.84 0.81   Normal
## 5 21.46 0.96      IFG
## 6 19.98 0.80   Normal

#analysis of data: mo ta cac bien theo gender

library(table1)

## Warning: package 'table1' was built under R version 3.5.3

## 
## Attaching package: 'table1'

## The following objects are masked from 'package:base':
## 
##     units, units<-

table1(~age+height+weight+bmi+waist+hip+whr+active+hypertension+diabetes|gender, data=db)

#do bien acitive va hypertension la bien phan loai nen dung factor truoc de biet la bien nhi phan

table1(~age+height+weight+bmi+waist+hip+whr+factor(active)+factor(hypertension) |diabetes, data=db)

#su khac biet giua bmi nam va nu

t.test(db$bmi~db$gender)

## 
##  Welch Two Sample t-test
## 
## data:  db$bmi by db$gender
## t = 0.66677, df = 2034.2, p-value = 0.505
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
##  -0.1749957  0.3552877
## sample estimates:
## mean in group Female   mean in group Male 
##             23.98515             23.89500

# ket luan: khong co su khac biet giua hai gioi
# su khac biet ve tieu duong giua nam va nu khong, do tieu duong co 03 gia tri nen dung t-test se khong nen. vi vay tao bien moi co hai gia tri la yes va no
db$diab=ifelse(db$diabetes=="Yes", "Yes", "No")
#dem so tieu duong cho moi nhom nam va nu
tab=table(db$diab, db$gender)
tab

##      
##       Female Male
##   No    2014  909
##   Yes    147   95

# kiem dinh z ve su khac biet giua 2 ti le tieu duong giua nam va nu bang kiem dinh chi binh phuong= prop.test()
prop.test(tab)

## 
##  2-sample test for equality of proportions with continuity
##  correction
## 
## data:  tab
## X-squared = 6.496, df = 1, p-value = 0.01081
## alternative hypothesis: two.sided
## 95 percent confidence interval:
##  0.01557132 0.14758891
## sample estimates:
##    prop 1    prop 2 
## 0.6890181 0.6074380

# Kl: co su khac biet ve tieu duong giua nam va nu
tab=table(db$gender, db$diab)
tab

##         
##            No  Yes
##   Female 2014  147
##   Male    909   95

prop.test(tab)

## 
##  2-sample test for equality of proportions with continuity
##  correction
## 
## data:  tab
## X-squared = 6.496, df = 1, p-value = 0.01081
## alternative hypothesis: two.sided
## 95 percent confidence interval:
##  0.004880546 0.048314356
## sample estimates:
##    prop 1    prop 2 
## 0.9319759 0.9053785

#su khac biet bmi giua 3 nhom diabetes: do diabetes co 3 nhom nen phan tich phuong sai: ANOVA

#aov: so sanh giua 3 nhom ve bmi
m=aov(bmi~diabetes, data=db)
summary(m)

##               Df Sum Sq Mean Sq F value  Pr(>F)   
## diabetes       2    151   75.27   5.851 0.00291 **
## Residuals   3153  40567   12.87                   
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 9 observations deleted due to missingness

#TukeyHSD : cho biet su khac biet giua tung nhom voi nhau
TukeyHSD(m)

##   Tukey multiple comparisons of means
##     95% family-wise confidence level
## 
## Fit: aov(formula = bmi ~ diabetes, data = db)
## 
## $diabetes
##                  diff        lwr          upr     p adj
## Normal-IFG -0.8236127 -1.3882150 -0.259010503 0.0018313
## Yes-IFG    -0.7643882 -1.5297871  0.001010667 0.0503925
## Yes-Normal  0.0592245 -0.5064508  0.624899835 0.9673231

# tao ra mot data cho nu
women=subset(db, gender=="Female")
dim(women)

## [1] 2161   15

m=aov(bmi~diabetes, data=women)
summary(m)

##               Df Sum Sq Mean Sq F value Pr(>F)
## diabetes       2     55   27.50   2.076  0.126
## Residuals   2150  28475   13.24               
## 8 observations deleted due to missingness

TukeyHSD(m)

##   Tukey multiple comparisons of means
##     95% family-wise confidence level
## 
## Fit: aov(formula = bmi ~ diabetes, data = women)
## 
## $diabetes
##                  diff       lwr       upr     p adj
## Normal-IFG -0.5165746 -1.228157 0.1950074 0.2044133
## Yes-IFG    -0.8256674 -1.808504 0.1571697 0.1198536
## Yes-Normal -0.3090928 -1.042806 0.4246202 0.5845178

# ve bieu do tuong quan da chieu cac bien lien tuc trong dataset
#tao ra mot dataframe moi gom cac bien lien tuc
dat=db[,c("age", "height", "weight","bmi", "waist", "hip", "whr")]
library(psych)
pairs.panels(dat)