Việc 1: Phân tích mô tả
1.1. Đọc dữ liệu “Obesity data.csv” vào R và gọi dữ liệu là
“ob”
ob = read.csv("D:\\nckh\\Obesity data.csv")
1.2. Mô tả đặc điểm tuổi (age), giới tính (gender), cân nặng
(weight), chiều cao (height), tỉ trọng mỡ (pcfat), tiền căn bệnh cao
huyết áp (hypertension)và tiền căn bệnh tiểu đường (diabetes):
summary(ob)
## id gender height weight
## Min. : 1.0 Length:1217 Min. :136.0 Min. :34.00
## 1st Qu.: 309.0 Class :character 1st Qu.:151.0 1st Qu.:49.00
## Median : 615.0 Mode :character Median :155.0 Median :54.00
## Mean : 614.5 Mean :156.7 Mean :55.14
## 3rd Qu.: 921.0 3rd Qu.:162.0 3rd Qu.:61.00
## Max. :1227.0 Max. :185.0 Max. :95.00
## bmi age WBBMC wbbmd fat
## Min. :14.5 Min. :13.00 Min. : 695 Min. :0.650 Min. : 4277
## 1st Qu.:20.2 1st Qu.:35.00 1st Qu.:1481 1st Qu.:0.930 1st Qu.:13768
## Median :22.2 Median :48.00 Median :1707 Median :1.010 Median :16955
## Mean :22.4 Mean :47.15 Mean :1725 Mean :1.009 Mean :17288
## 3rd Qu.:24.3 3rd Qu.:58.00 3rd Qu.:1945 3rd Qu.:1.090 3rd Qu.:20325
## Max. :37.1 Max. :88.00 Max. :3040 Max. :1.350 Max. :40825
## lean pcfat hypertension diabetes
## Min. :19136 Min. : 9.2 Min. :0.000 Min. :0.0000
## 1st Qu.:30325 1st Qu.:27.0 1st Qu.:0.000 1st Qu.:0.0000
## Median :33577 Median :32.4 Median :1.000 Median :0.0000
## Mean :35463 Mean :31.6 Mean :0.507 Mean :0.1109
## 3rd Qu.:39761 3rd Qu.:36.8 3rd Qu.:1.000 3rd Qu.:0.0000
## Max. :63059 Max. :48.4 Max. :1.000 Max. :1.0000
1.3. Nhận xét về kết quả của tiền căn bệnh cao huyết áp và tiểu
đường
table(ob$hypertension)
##
## 0 1
## 600 617
a.Tính tỷ lệ % cho bệnh cao huyết áp
prop.table(table(ob$hypertension)) * 100
##
## 0 1
## 49.30156 50.69844
b. Đếm số lượng cho bệnh tiểu đường
table(ob$diabetes)
##
## 0 1
## 1082 135
c. Tính tỷ lệ % cho bệnh tiểu đường
prop.table(table(ob$diabetes)) * 100
##
## 0 1
## 88.90715 11.09285
1.4. Trình bày kết quả trung vị (Q1, Q3) thay vi trung vị (min, max)
cho biến liên tục.
nam <- subset(ob, gender == "M")
nu <- subset(ob, gender == "F")
summary(nam)
## id gender height weight
## Min. : 2.0 Length:355 Min. :146.0 Min. :38.00
## 1st Qu.: 288.5 Class :character 1st Qu.:160.0 1st Qu.:55.00
## Median : 531.0 Mode :character Median :165.0 Median :62.00
## Mean : 558.0 Mean :165.1 Mean :62.02
## 3rd Qu.: 844.5 3rd Qu.:169.0 3rd Qu.:68.00
## Max. :1219.0 Max. :185.0 Max. :95.00
## bmi age WBBMC wbbmd fat
## Min. :14.50 Min. :13.0 Min. :1194 Min. :0.78 Min. : 4277
## 1st Qu.:20.80 1st Qu.:24.0 1st Qu.:1805 1st Qu.:0.99 1st Qu.:11428
## Median :22.50 Median :44.0 Median :2026 Median :1.06 Median :15149
## Mean :22.73 Mean :43.7 Mean :2030 Mean :1.06 Mean :14978
## 3rd Qu.:24.85 3rd Qu.:56.0 3rd Qu.:2252 3rd Qu.:1.13 3rd Qu.:18182
## Max. :34.70 Max. :88.0 Max. :3040 Max. :1.34 Max. :29944
## lean pcfat hypertension diabetes
## Min. :28587 Min. : 9.20 Min. :0.0000 Min. :0.00000
## 1st Qu.:40324 1st Qu.:20.35 1st Qu.:0.0000 1st Qu.:0.00000
## Median :43391 Median :24.60 Median :1.0000 Median :0.00000
## Mean :43762 Mean :24.16 Mean :0.5211 Mean :0.09296
## 3rd Qu.:47563 3rd Qu.:28.00 3rd Qu.:1.0000 3rd Qu.:0.00000
## Max. :63059 Max. :39.00 Max. :1.0000 Max. :1.00000
summary(nu)
## id gender height weight
## Min. : 1.0 Length:862 Min. :136.0 Min. :34.00
## 1st Qu.: 318.0 Class :character 1st Qu.:150.0 1st Qu.:47.00
## Median : 646.5 Mode :character Median :153.0 Median :51.00
## Mean : 637.8 Mean :153.3 Mean :52.31
## 3rd Qu.: 967.8 3rd Qu.:157.0 3rd Qu.:57.00
## Max. :1227.0 Max. :170.0 Max. :95.00
## bmi age WBBMC wbbmd
## Min. :15.20 Min. :14.00 Min. : 695 Min. :0.6500
## 1st Qu.:20.10 1st Qu.:39.00 1st Qu.:1408 1st Qu.:0.9100
## Median :22.10 Median :49.00 Median :1614 Median :0.9900
## Mean :22.26 Mean :48.57 Mean :1599 Mean :0.9876
## 3rd Qu.:24.10 3rd Qu.:59.00 3rd Qu.:1802 3rd Qu.:1.0700
## Max. :37.10 Max. :85.00 Max. :2663 Max. :1.3500
## fat lean pcfat hypertension
## Min. : 6222 Min. :19136 Min. :14.60 Min. :0.0000
## 1st Qu.:14795 1st Qu.:29336 1st Qu.:31.50 1st Qu.:0.0000
## Median :17695 Median :31504 Median :34.70 Median :1.0000
## Mean :18240 Mean :32045 Mean :34.67 Mean :0.5012
## 3rd Qu.:21120 3rd Qu.:34464 3rd Qu.:38.30 3rd Qu.:1.0000
## Max. :40825 Max. :53350 Max. :48.40 Max. :1.0000
## diabetes
## Min. :0.0000
## 1st Qu.:0.0000
## Median :0.0000
## Mean :0.1183
## 3rd Qu.:0.0000
## Max. :1.0000
1.5. Mô tả đặc điểm tuổi (age), cân nặng (weight), chiều cao
(height), tỉ trọng mỡ (pcfat) và tiền căn bệnh cao huyết áp
(hypertension) theo giới tính (gender)
prop.table(table(nam$hypertension))*100
##
## 0 1
## 47.88732 52.11268
prop.table(table(nu$hypertension))*100
##
## 0 1
## 49.88399 50.11601
1.6. Đánh giá xem đặc điểm nào là khác biệt đáng kể (significant
difference) giữa nam và nữ
t.test(nam$age, nu$age)
##
## Welch Two Sample t-test
##
## data: nam$age and nu$age
## t = -4.2608, df = 587.49, p-value = 2.372e-05
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
## -7.117272 -2.626083
## sample estimates:
## mean of x mean of y
## 43.70141 48.57309
wilcox.test(nam$weight, nu$weight)
##
## Wilcoxon rank sum test with continuity correction
##
## data: nam$weight and nu$weight
## W = 241840, p-value < 2.2e-16
## alternative hypothesis: true location shift is not equal to 0
bang_so_sanh <- table(ob$gender, ob$hypertension)
print(bang_so_sanh)
##
## 0 1
## F 430 432
## M 170 185
chisq.test(bang_so_sanh)
##
## Pearson's Chi-squared test with Yates' continuity correction
##
## data: bang_so_sanh
## X-squared = 0.32515, df = 1, p-value = 0.5685
việc 2: Phân tích khác biệt giữa 2 nhóm
2.1. Giả sử bạn có dữ liệu về tải trọng của 2 nhóm A và B như
sau
Nhom_A <- c(14, 4, 10, 6, 3, 11, 12)
Nhom_B <- c(16, 17, 13, 12, 7, 16, 11, 8, 7)
2.2. Kiểm tra tải trọng có tuân theo phân bố chuẩn
a. Kiểm tra phân bố chuẩn cho Nhóm A
shapiro.test(Nhom_A)
##
## Shapiro-Wilk normality test
##
## data: Nhom_A
## W = 0.92541, p-value = 0.5126
b. Kiểm tra phân bố chuẩn cho Nhóm B
shapiro.test(Nhom_B)
##
## Shapiro-Wilk normality test
##
## data: Nhom_B
## W = 0.89641, p-value = 0.2319
2.3. Mô tả đặc điểm tải trọng giữa hai nhóm
summary(Nhom_A)
## Min. 1st Qu. Median Mean 3rd Qu. Max.
## 3.000 5.000 10.000 8.571 11.500 14.000
summary(Nhom_B)
## Min. 1st Qu. Median Mean 3rd Qu. Max.
## 7.00 8.00 12.00 11.89 16.00 17.00
2.4. Thực hiện kiểm định t-test
t.test(Nhom_A, Nhom_B)
##
## Welch Two Sample t-test
##
## data: Nhom_A and Nhom_B
## t = -1.6, df = 12.554, p-value = 0.1345
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
## -7.813114 1.178194
## sample estimates:
## mean of x mean of y
## 8.571429 11.888889
2.5. Thực hiện bootstrap để đánh giá khác biệt trung bình
set.seed(123)
bootstrap_diff_mean <- replicate(1000, {
sample_A <- sample(Nhom_A, replace = TRUE)
sample_B <- sample(Nhom_B, replace = TRUE)
mean(sample_A) - mean(sample_B)
})
quantile(bootstrap_diff_mean, c(0.025, 0.5, 0.975))
## 2.5% 50% 97.5%
## -7.0952381 -3.3174603 0.4924603
hist(bootstrap_diff_mean, main = "Bootstrap Difference in Means", xlab = "Mean Difference")
2.6. Bootstrap so sánh trung vị
set.seed(123)
bootstrap_diff_median <- replicate(1000, {
sample_A <- sample(Nhom_A, replace = TRUE)
sample_B <- sample(Nhom_B, replace = TRUE)
median(sample_A) - median(sample_B)
})
quantile(bootstrap_diff_median, c(0.025, 0.5, 0.975))
## 2.5% 50% 97.5%
## -10 -2 4
hist(bootstrap_diff_median, main = "Bootstrap Difference in Medians", xlab = "Median Difference")
