Bai tap ngay 1 - BV Thong Nhat
Tuan Nguyen
2025-12-29
R Markdown
Đây là một phiên thực hành R trên R Markdown. File này sẽ đọc dữ liệu vào R và thực hiện một số phân tích căn bản.
Bắt đầu đọc dữ liệu vào R
library(Hmisc)##
## Attaching package: 'Hmisc'## The following objects are masked from 'package:base':
##
## format.pval, unitsdf = read.csv("/Users/121493/Dropbox/_Conferences and Workshops/BV Thong Nhat/Tháng 12-2025/Dataset/Stroke Data.csv")
# Liệt kê 6 dòng đầu
head(df, 10)## id gender age hypertension heart.disease ever.married work.type
## 1 67 Female 17 0 0 No Private
## 2 77 Female 13 0 0 No children
## 3 84 Male 55 0 0 Yes Private
## 4 91 Female 42 0 0 No Private
## 5 99 Female 31 0 0 No Private
## 6 121 Female 38 0 0 Yes Private
## 7 129 Female 24 0 0 No Private
## 8 132 Female 80 0 0 Yes Govt_job
## 9 156 Female 33 0 0 Yes Private
## 10 163 Female 20 0 0 No Private
## Residence.type glucose.level bmi smoking stroke
## 1 Urban 92.97 NA formerly smoked 0
## 2 Rural 85.81 18.6 Unknown 0
## 3 Urban 89.17 31.5 never smoked 0
## 4 Urban 98.53 18.5 never smoked 0
## 5 Urban 108.89 52.3 Unknown 0
## 6 Urban 91.44 NA Unknown 0
## 7 Urban 97.55 26.2 never smoked 0
## 8 Urban 84.86 NA Unknown 0
## 9 Rural 86.97 42.2 never smoked 0
## 10 Rural 94.67 28.8 Unknown 0# Mã hoá
library(lessR)##
## lessR 4.3.9 feedback: gerbing@pdx.edu
## --------------------------------------------------------------
## > d <- Read("") Read text, Excel, SPSS, SAS, or R data file
## d is default data frame, data= in analysis routines optional
##
## Many examples of reading, writing, and manipulating data,
## graphics, testing means and proportions, regression, factor analysis,
## customization, and descriptive statistics from pivot tables
## Enter: browseVignettes("lessR")
##
## View lessR updates, now including time series forecasting
## Enter: news(package="lessR")
##
## Interactive data analysis
## Enter: interact()##
## Attaching package: 'lessR'## The following objects are masked from 'package:Hmisc':
##
## label, Merge## The following object is masked from 'package:base':
##
## sort_bydf$stroke <- factor(df$stroke, levels = c(0, 1), labels = c("No", "Yes"))
# You may also want to ensure other categorical variables are factors
# Adjust these based on your actual data structure
df$gender <- as.factor(df$gender)
df$hypertension <- factor(df$hypertension, levels = c(0, 1), labels = c("No", "Yes"))
df$heart.disease <- factor(df$heart.disease, levels = c(0, 1), labels = c("No", "Yes"))
df$ever.married <- as.factor(df$ever.married)
df$work.type <- as.factor(df$work.type)
df$Residence.type <- as.factor(df$Residence.type)
df$smoking <- as.factor(df$smoking)
# Add labels for better presentation
label(df$gender) <- "Gender"
label(df$age) <- "Age (years)"
label(df$hypertension) <- "Hypertension"
label(df$heart.disease) <- "Heart Disease"
label(df$ever.married) <- "Ever Married"
label(df$work.type) <- "Work Type"
label(df$Residence.type) <- "Residence Type"
label(df$glucose.level) <- "Glucose Level (mg/dL)"
label(df$bmi) <- "BMI (kg/m²)"
label(df$smoking) <- "Smoking Status"
head(df)## id gender age hypertension heart.disease ever.married work.type
## 1 67 Female 17 No No No Private
## 2 77 Female 13 No No No children
## 3 84 Male 55 No No Yes Private
## 4 91 Female 42 No No No Private
## 5 99 Female 31 No No No Private
## 6 121 Female 38 No No Yes Private
## Residence.type glucose.level bmi smoking stroke
## 1 Urban 92.97 NA formerly smoked No
## 2 Rural 85.81 18.6 Unknown No
## 3 Urban 89.17 31.5 never smoked No
## 4 Urban 98.53 18.5 never smoked No
## 5 Urban 108.89 52.3 Unknown No
## 6 Urban 91.44 NA Unknown NoPhân tích mô tả dùng table1
library(table1)##
## Attaching package: 'table1'## The following object is masked from 'package:lessR':
##
## label## The following objects are masked from 'package:Hmisc':
##
## label, label<-, units## The following objects are masked from 'package:base':
##
## units, units<-# Phân tích toàn bộ dữ liệu
table1(~age + gender + hypertension + heart.disease + ever.married + work.type + Residence.type + glucose.level + bmi + smoking, data=df)| Overall (N=5110) |
|
|---|---|
| Age (years) | |
| Mean (SD) | 43.2 (22.6) |
| Median [Min, Max] | 45.0 [0.0800, 82.0] |
| Gender | |
| Female | 2994 (58.6%) |
| Male | 2116 (41.4%) |
| Hypertension | |
| No | 4612 (90.3%) |
| Yes | 498 (9.7%) |
| Heart Disease | |
| No | 4834 (94.6%) |
| Yes | 276 (5.4%) |
| Ever Married | |
| No | 1757 (34.4%) |
| Yes | 3353 (65.6%) |
| Work Type | |
| children | 687 (13.4%) |
| Govt_job | 657 (12.9%) |
| Never_worked | 22 (0.4%) |
| Private | 2925 (57.2%) |
| Self-employed | 819 (16.0%) |
| Residence Type | |
| Rural | 2514 (49.2%) |
| Urban | 2596 (50.8%) |
| Glucose Level (mg/dL) | |
| Mean (SD) | 106 (45.3) |
| Median [Min, Max] | 91.9 [55.1, 272] |
| BMI (kg/m²) | |
| Mean (SD) | 28.9 (7.85) |
| Median [Min, Max] | 28.1 [10.3, 97.6] |
| Missing | 201 (3.9%) |
| Smoking Status | |
| formerly smoked | 885 (17.3%) |
| never smoked | 1892 (37.0%) |
| smokes | 789 (15.4%) |
| Unknown | 1544 (30.2%) |
# Phân tích theo nhóm stroke
table1(~age + gender + hypertension + heart.disease + ever.married + work.type + Residence.type + glucose.level + bmi + smoking | stroke, data=df)| No (N=4861) |
Yes (N=249) |
Overall (N=5110) |
|
|---|---|---|---|
| Age (years) | |||
| Mean (SD) | 42.0 (22.3) | 67.7 (12.7) | 43.2 (22.6) |
| Median [Min, Max] | 43.0 [0.0800, 82.0] | 71.0 [1.32, 82.0] | 45.0 [0.0800, 82.0] |
| Gender | |||
| Female | 2853 (58.7%) | 141 (56.6%) | 2994 (58.6%) |
| Male | 2008 (41.3%) | 108 (43.4%) | 2116 (41.4%) |
| Hypertension | |||
| No | 4429 (91.1%) | 183 (73.5%) | 4612 (90.3%) |
| Yes | 432 (8.9%) | 66 (26.5%) | 498 (9.7%) |
| Heart Disease | |||
| No | 4632 (95.3%) | 202 (81.1%) | 4834 (94.6%) |
| Yes | 229 (4.7%) | 47 (18.9%) | 276 (5.4%) |
| Ever Married | |||
| No | 1728 (35.5%) | 29 (11.6%) | 1757 (34.4%) |
| Yes | 3133 (64.5%) | 220 (88.4%) | 3353 (65.6%) |
| Work Type | |||
| children | 685 (14.1%) | 2 (0.8%) | 687 (13.4%) |
| Govt_job | 624 (12.8%) | 33 (13.3%) | 657 (12.9%) |
| Never_worked | 22 (0.5%) | 0 (0%) | 22 (0.4%) |
| Private | 2776 (57.1%) | 149 (59.8%) | 2925 (57.2%) |
| Self-employed | 754 (15.5%) | 65 (26.1%) | 819 (16.0%) |
| Residence Type | |||
| Rural | 2400 (49.4%) | 114 (45.8%) | 2514 (49.2%) |
| Urban | 2461 (50.6%) | 135 (54.2%) | 2596 (50.8%) |
| Glucose Level (mg/dL) | |||
| Mean (SD) | 105 (43.8) | 133 (61.9) | 106 (45.3) |
| Median [Min, Max] | 91.5 [55.1, 268] | 105 [56.1, 272] | 91.9 [55.1, 272] |
| BMI (kg/m²) | |||
| Mean (SD) | 28.8 (7.91) | 30.5 (6.33) | 28.9 (7.85) |
| Median [Min, Max] | 28.0 [10.3, 97.6] | 29.7 [16.9, 56.6] | 28.1 [10.3, 97.6] |
| Missing | 161 (3.3%) | 40 (16.1%) | 201 (3.9%) |
| Smoking Status | |||
| formerly smoked | 815 (16.8%) | 70 (28.1%) | 885 (17.3%) |
| never smoked | 1802 (37.1%) | 90 (36.1%) | 1892 (37.0%) |
| smokes | 747 (15.4%) | 42 (16.9%) | 789 (15.4%) |
| Unknown | 1497 (30.8%) | 47 (18.9%) | 1544 (30.2%) |
Hiển thị dữ liệu dùng lessR
# Biểu đồ phân bố bmi
Histogram(bmi, data=df)
## >>> Suggestions
## bin_width: set the width of each bin
## bin_start: set the start of the first bin
## bin_end: set the end of the last bin
## Histogram(bmi, density=TRUE) # smoothed curve + histogram
## Plot(bmi) # Violin/Box/Scatterplot (VBS) plot
##
## --- bmi ---
##
## n miss mean sd min mdn max
## 4909 201 28.89 7.85 10.30 28.10 97.60
##
##
## --- Outliers --- from the box plot: 110
##
## Small Large
## ----- -----
## 97.6
## 92.0
## 78.0
## 71.9
## 66.8
## 64.8
## 64.4
## 63.3
## 61.6
## 61.2
## 60.9
## 60.9
## 60.2
## 59.7
## 58.1
## 57.9
## 57.7
## 57.5
##
## + 92 more outliers
##
##
## Bin Width: 5
## Number of Bins: 18
##
## Bin Midpnt Count Prop Cumul.c Cumul.p
## ---------------------------------------------------
## 10 > 15 12.5 44 0.01 44 0.01
## 15 > 20 17.5 493 0.10 537 0.11
## 20 > 25 22.5 1070 0.21 1607 0.31
## 25 > 30 27.5 1409 0.28 3016 0.59
## 30 > 35 32.5 985 0.19 4001 0.78
## 35 > 40 37.5 500 0.10 4501 0.88
## 40 > 45 42.5 253 0.05 4754 0.93
## 45 > 50 47.5 76 0.01 4830 0.95
## 50 > 55 52.5 46 0.01 4876 0.95
## 55 > 60 57.5 20 0.00 4896 0.96
## 60 > 65 62.5 8 0.00 4904 0.96
## 65 > 70 67.5 1 0.00 4905 0.96
## 70 > 75 72.5 1 0.00 4906 0.96
## 75 > 80 77.5 1 0.00 4907 0.96
## 80 > 85 82.5 0 0.00 4907 0.96
## 85 > 90 87.5 0 0.00 4907 0.96
## 90 > 95 92.5 1 0.00 4908 0.96
## 95 > 100 97.5 1 0.00 4909 0.96Histogram(bmi, fill="blue", xlab="Body mass index (g)", ylab="Frequency", data=df)
## >>> Suggestions
## bin_width: set the width of each bin
## bin_start: set the start of the first bin
## bin_end: set the end of the last bin
## Histogram(bmi, density=TRUE) # smoothed curve + histogram
## Plot(bmi) # Violin/Box/Scatterplot (VBS) plot
##
## --- bmi ---
##
## n miss mean sd min mdn max
## 4909 201 28.89 7.85 10.30 28.10 97.60
##
##
## --- Outliers --- from the box plot: 110
##
## Small Large
## ----- -----
## 97.6
## 92.0
## 78.0
## 71.9
## 66.8
## 64.8
## 64.4
## 63.3
## 61.6
## 61.2
## 60.9
## 60.9
## 60.2
## 59.7
## 58.1
## 57.9
## 57.7
## 57.5
##
## + 92 more outliers
##
##
## Bin Width: 5
## Number of Bins: 18
##
## Bin Midpnt Count Prop Cumul.c Cumul.p
## ---------------------------------------------------
## 10 > 15 12.5 44 0.01 44 0.01
## 15 > 20 17.5 493 0.10 537 0.11
## 20 > 25 22.5 1070 0.21 1607 0.31
## 25 > 30 27.5 1409 0.28 3016 0.59
## 30 > 35 32.5 985 0.19 4001 0.78
## 35 > 40 37.5 500 0.10 4501 0.88
## 40 > 45 42.5 253 0.05 4754 0.93
## 45 > 50 47.5 76 0.01 4830 0.95
## 50 > 55 52.5 46 0.01 4876 0.95
## 55 > 60 57.5 20 0.00 4896 0.96
## 60 > 65 62.5 8 0.00 4904 0.96
## 65 > 70 67.5 1 0.00 4905 0.96
## 70 > 75 72.5 1 0.00 4906 0.96
## 75 > 80 77.5 1 0.00 4907 0.96
## 80 > 85 82.5 0 0.00 4907 0.96
## 85 > 90 87.5 0 0.00 4907 0.96
## 90 > 95 92.5 1 0.00 4908 0.96
## 95 > 100 97.5 1 0.00 4909 0.96# Biểu đồ thanh
BarChart(work.type, data=df)
## >>> Suggestions
## BarChart(work.type, horiz=TRUE) # horizontal bar chart
## BarChart(work.type, fill="reds") # red bars of varying lightness
## PieChart(work.type) # doughnut (ring) chart
## Plot(work.type) # bubble plot
## Plot(work.type, stat="count") # lollipop plot
##
## --- work.type ---
##
## Missing Values: 0
##
## work.type Count Prop
## ----------------------------
## children 687 0.134
## Govt_job 657 0.129
## Never_worked 22 0.004
## Private 2925 0.572
## Self-employed 819 0.160
## ----------------------------
## Total 5110 1.000
##
## Chi-squared test of null hypothesis of equal probabilities
## Chisq = 4802.415, df = 4, p-value = 0.000# Biểu đồ tương quan
Plot(age, bmi, fit="lm", data=df)
##
## >>> Suggestions or enter: style(suggest=FALSE)
## Plot(age, bmi, enhance=TRUE) # many options
## Plot(age, bmi, color="red") # exterior edge color of points
## Plot(age, bmi, MD_cut=6) # Mahalanobis distance from center > 6 is an outlier
##
##
## >>> Pearson's product-moment correlation
##
## Number of paired values with neither missing, n = 4909
## Sample Correlation of age and bmi: r = 0.333
##
## Hypothesis Test of 0 Correlation: t = 24.772, df = 4907, p-value = 0.000
## 95% Confidence Interval for Correlation: 0.308 to 0.358
##
##
## Line: b0 = 23.92 b1 = 0.12 Fit: MSE = 54.841 Rsq = 0.111
## Mô hình hồi qui logistic
fit = Logit(stroke ~ gender + age, data=df)##
## >>> Note: gender is not a numeric variable.
## Indicator variables are created and analyzed.
##
## Response Variable: stroke
## Predictor Variable 1: genderMale
## Predictor Variable 2: age
##
## Number of cases (rows) of data: 5110
## Number of cases retained for analysis: 5110
##
##
## BASIC ANALYSIS
##
## -- Estimated Model of stroke for the Logit of Reference Group Membership
##
## Estimate Std Err z-value p-value Lower 95% Upper 95%
## (Intercept) -7.2834 0.3418 -21.308 0.000 -7.9533 -6.6134
## genderMale 0.1153 0.1374 0.839 0.401 -0.1539 0.3845
## age 0.0748 0.0049 15.167 0.000 0.0651 0.0844
##
##
## -- Odds Ratios and Confidence Intervals
##
## Odds Ratio Lower 95% Upper 95%
## (Intercept) 0.0007 0.0004 0.0013
## genderMale 1.1222 0.8573 1.4689
## age 1.0776 1.0673 1.0881
##
##
## -- Model Fit
##
## Null deviance: 1990.373 on 5109 degrees of freedom
## Residual deviance: 1615.595 on 5107 degrees of freedom
##
## AIC: 1621.595
##
## Number of iterations to convergence: 7
##
##
## Collinearity
##
## Tolerance VIF
## genderMale 0.999 1.001
## age 0.999 1.001
##
## ANALYSIS OF RESIDUALS AND INFLUENCE
## Data, Fitted, Residual, Studentized Residual, Dffits, Cook's Distance
## [sorted by Cook's Distance]
## [res_rows = 20 out of 5110 cases (rows) of data]
## --------------------------------------------------------------------
## genderMale age stroke P(Y=1) residual rstudent dffits cooks
## 4897 0 1 Yes 0.0007396 0.9993 3.812 0.06215 0.037826
## 3451 0 14 Yes 0.0019529 0.9980 3.543 0.07678 0.025275
## 2764 0 32 Yes 0.0074615 0.9925 3.136 0.09328 0.012387
## 2190 0 38 Yes 0.0116372 0.9884 2.989 0.09620 0.009255
## 2205 0 38 Yes 0.0116372 0.9884 2.989 0.09620 0.009255
## 2159 0 39 Yes 0.0125294 0.9875 2.964 0.09651 0.008790
## 2349 0 39 Yes 0.0125294 0.9875 2.964 0.09651 0.008790
## 2507 0 39 Yes 0.0125294 0.9875 2.964 0.09651 0.008790
## 2345 1 42 Yes 0.0175082 0.9825 2.849 0.10540 0.008086
## 2922 1 43 Yes 0.0188421 0.9812 2.822 0.10574 0.007692
## 3818 1 45 Yes 0.0218153 0.9782 2.770 0.10630 0.006953
## 1387 0 45 Yes 0.0194856 0.9805 2.810 0.09724 0.006341
## 2848 0 45 Yes 0.0194856 0.9805 2.810 0.09724 0.006341
## 1051 1 47 Yes 0.0252456 0.9748 2.716 0.10675 0.006278
## 3827 0 46 Yes 0.0209669 0.9790 2.783 0.09719 0.005989
## 1029 1 48 Yes 0.0271526 0.9728 2.689 0.10693 0.005964
## 659 1 49 Yes 0.0291993 0.9708 2.662 0.10710 0.005667
## 4688 0 48 Yes 0.0242671 0.9757 2.730 0.09696 0.005336
## 4636 1 51 Yes 0.0337507 0.9662 2.606 0.10742 0.005119
## 4227 0 49 Yes 0.0261021 0.9739 2.703 0.09679 0.005033
##
##
## PREDICTION
##
## Probability threshold for classification Yes: 0.5
##
## 0: No
## 1: Yes
##
## Data, Fitted Values, Standard Errors
## [sorted by fitted value]
## [pred_all=TRUE to see all intervals displayed]
## --------------------------------------------------------------------
## genderMale age stroke label fitted std.err
## 136 0 0 No 0 0.0006864 0.0002345
## 556 0 0 No 0 0.0006864 0.0002345
## 1135 0 0 No 0 0.0006864 0.0002345
## 1597 0 0 No 0 0.0006864 0.0002345
##
## ... for the rows of data where fitted is close to 0.5 ...
##
## genderMale age stroke label fitted std.err
## 4624 1 81 No 0 0.2477 0.02353
## 4885 1 81 No 0 0.2477 0.02353
## 337 1 82 No 0 0.2619 0.02497
## 546 1 82 No 0 0.2619 0.02497
## 1367 1 82 No 0 0.2619 0.02497
##
## ... for the last 4 rows of sorted data ...
##
## genderMale age stroke label fitted std.err
## 4533 1 82 Yes 0 0.2619 0.02497
## 4799 1 82 No 0 0.2619 0.02497
## 4921 1 82 No 0 0.2619 0.02497
## 4925 1 82 No 0 0.2619 0.02497
## --------------------------------------------------------------------
##
##
## ----------------------------
## Specified confusion matrices
## ----------------------------
##
## Probability threshold for predicting Yes: 0.5
##
## Baseline Predicted
## ---------------------------------------------------
## Total %Tot 0 1 %Correct
## ---------------------------------------------------
## 1 249 4.9 249 0 0.0
## stroke 0 4861 95.1 4861 0 100.0
## ---------------------------------------------------
## Total 5110 95.1
##
## Accuracy: 95.13
## Sensitivity: 0.00
## Precision: NaN