Data Wrangling (Data Preprocessing)

Practical Assessment 1

Setup: Load packages

library(readr)
library(dplyr) 
library(kableExtra) #enhance and customize tables

Attaching package: ‘kableExtra’

The following object is masked from ‘package:dplyr’:

    group_rows

library(magrittr) #provides pipe operator

Attaching package: ‘magrittr’

The following object is masked from ‘package:purrr’:

    set_names

The following object is masked from ‘package:tidyr’:

    extract

Group Information

Group Information

Student name	Student number	Percentage of contribution
Chamudi Abeysinghe	s4150303	100%

Description of the Data

AKSHAY KUMAR (2025) states that ‘This datas real-world trends in children’s screen time usage. It includes data on educational, recreational, and total screen time for children aged 5 to 15 years, with breakdowns by gender (Male, Female, Other/Prefer not to say) and day type (Weekday, Weekend)’.

Available at:https://www.kaggle.com/datasets/ak0212/average-daily-screen-time-for-children

Variable Description:

1. Age: Children age between 5-15 years. Helps analyze how screen time habits change as children grow older.

2. Gender: The gender of the child (Male,Female,Other/Prefer not to say). Allows exploration of gender-based differences in screen time behavior.

3. Screen Type Time: The category of screen use being measured. Includes data on educational, recreational, and total screen time.

4. Day Type: Indicates whether the data was collected for a weekday or weekend. Helps analyze how screen time varies between school days and off days.

5. Average Screen Time (hours): The average number of hours spent on screens per day for the given category.

6. Sample Size: Number of survey respondents represented by the data point. Larger sample sizes usually mean more reliable data.

Import Data

getwd()

[1] "/Users/chamudi/Desktop/RMIT/1st Year 1st Semester/Data Wrangling/Assignments/Updated Assignment 1"

screen_time_ds <- read.csv('Datasets/screen_time.csv')

head(screen_time_ds)

str(screen_time_ds) #to show the structure of the data (types, columns, etc.)

'data.frame':   198 obs. of  6 variables:
 $ Age                        : int  5 5 5 5 5 5 5 5 5 5 ...
 $ Gender                     : chr  "Male" "Male" "Male" "Male" ...
 $ Screen.Time.Type           : chr  "Educational" "Recreational" "Total" "Educational" ...
 $ Day.Type                   : chr  "Weekday" "Weekday" "Weekday" "Weekend" ...
 $ Average.Screen.Time..hours.: num  0.44 1.11 1.55 0.5 1.44 1.93 0.49 0.96 1.45 0.5 ...
 $ Sample.Size                : int  500 500 500 500 500 500 500 500 500 500 ...

summary(screen_time_ds) #to get descriptive statistics

      Age        Gender          Screen.Time.Type     Day.Type        
 Min.   : 5   Length:198         Length:198         Length:198        
 1st Qu.: 7   Class :character   Class :character   Class :character  
 Median :10   Mode  :character   Mode  :character   Mode  :character  
 Mean   :10                                                           
 3rd Qu.:13                                                           
 Max.   :15                                                           
 Average.Screen.Time..hours.  Sample.Size 
 Min.   :0.440               Min.   :300  
 1st Qu.:1.403               1st Qu.:340  
 Median :2.490               Median :400  
 Mean   :2.993               Mean   :400  
 3rd Qu.:4.397               3rd Qu.:460  
 Max.   :8.190               Max.   :500  

dim(screen_time_ds) #number of rows and columns

[1] 198   6

names(screen_time_ds) #column names

[1] "Age"                         "Gender"                     
[3] "Screen.Time.Type"            "Day.Type"                   
[5] "Average.Screen.Time..hours." "Sample.Size"                

First I downloaded a csv data set from Kaggle. Then loaded the libraries. Then I made sure which working directory I am working in. And then using the read.csv package, I imported the data set into an object called ‘screen_time_ds’. Next I loaded the first few rows of the data frame to confirm that the data was read correctly.

I have used other outputs, beside head().

In R, when we use read.csv() and assign it to a object (like screen_time_ds), we are storing the data set in a data frame. In R, a data frame is simply an object that holds tabular data, and by default, read.csv() returns a data frame.

Inspect and Understand

#Check the dimensions of the data frame.
dim(screen_time_ds) 

[1] 198   6

#Check the column names in the data frame, rename them if required.
colnames(screen_time_ds)

[1] "Age"                         "Gender"                     
[3] "Screen.Time.Type"            "Day.Type"                   
[5] "Average.Screen.Time..hours." "Sample.Size"                

screen_time_ds <- screen_time_ds %>%
  rename(
    Screen_Type = Screen.Time.Type,
    Day_Type = Day.Type,
    Avg_Screen_Time = Average.Screen.Time..hours.,
    Sample_Size = Sample.Size
  )

#Summaries the types of variables

sapply(screen_time_ds, class) #https://stackoverflow.com/questions/21125222/determine-the-data-types-of-a-data-frames-columns

            Age          Gender     Screen_Type        Day_Type Avg_Screen_Time 
      "integer"     "character"     "character"     "character"       "numeric" 
    Sample_Size 
      "integer" 

str(screen_time_ds)

'data.frame':   198 obs. of  6 variables:
 $ Age            : int  5 5 5 5 5 5 5 5 5 5 ...
 $ Gender         : chr  "Male" "Male" "Male" "Male" ...
 $ Screen_Type    : chr  "Educational" "Recreational" "Total" "Educational" ...
 $ Day_Type       : chr  "Weekday" "Weekday" "Weekday" "Weekend" ...
 $ Avg_Screen_Time: num  0.44 1.11 1.55 0.5 1.44 1.93 0.49 0.96 1.45 0.5 ...
 $ Sample_Size    : int  500 500 500 500 500 500 500 500 500 500 ...

#Check the levels of factor variables, rename/rearrange them if required.
levels(screen_time_ds$Age)

NULL

levels(screen_time_ds$Gender)

NULL

levels(screen_time_ds$Screen_Type)

NULL

levels(screen_time_ds$Day_Type)

NULL

levels(screen_time_ds$Avg_Screen_Time)

NULL

levels(screen_time_ds$Sample_Size)

NULL

#convert 'Gender' and 'Day_Type' character column into factor variable
screen_time_ds$Gender <- as.factor(screen_time_ds$Gender)
screen_time_ds$Day_Type <- as.factor(screen_time_ds$Day_Type)

#Check the levels of factor variables after converting
levels(screen_time_ds$Gender)

[1] "Female"                  "Male"                    "Other/Prefer not to say"

levels(screen_time_ds$Day_Type)

[1] "Weekday" "Weekend"

str(screen_time_ds)

'data.frame':   198 obs. of  6 variables:
 $ Age            : int  5 5 5 5 5 5 5 5 5 5 ...
 $ Gender         : Factor w/ 3 levels "Female","Male",..: 2 2 2 2 2 2 1 1 1 1 ...
 $ Screen_Type    : chr  "Educational" "Recreational" "Total" "Educational" ...
 $ Day_Type       : Factor w/ 2 levels "Weekday","Weekend": 1 1 1 2 2 2 1 1 1 2 ...
 $ Avg_Screen_Time: num  0.44 1.11 1.55 0.5 1.44 1.93 0.49 0.96 1.45 0.5 ...
 $ Sample_Size    : int  500 500 500 500 500 500 500 500 500 500 ...

First, I checked the dimensions of the data frame. I used the dim() function and it gave the output 198 observations and 6 variables.

Secondly, I checked the column names in the data frame. I used the colnames() function to get the column names in the data frame. I decided to rename “Screen.Time.Type”, “Day.Type”, “Average.Screen.Time..hours.” and “Sample.Size” column names (replacing long and awkward column names with cleaner versions). Therefore I used rename() function. I saved new column names into the same object where the data set is saved before. I passed data frame using pipe operator for the rename().

Next, I got the summary of the data set. I used the ‘sapply(dataset,class)’ function. I used stack overflow to find a way other that str() function to do this task. Then I used str() and gave the structure summary of our data set, including the class/type of each column. All data types are correct in this data set. Therefore, there is no need of conversion.

If there is a incorrect type, it is given as integer and I want to make it as numeric, so, I use, screen_time_ds\('columnName' <- as.Numeric(screen_time_ds\)’columnName’) function.

Finally, Columns like Gender or Day_Type has a small set of categories, storing them as factors is more appropriate than as strings. So, I decided to change them into factor variables. I didn’t label those factor variables newly and I kept them as it is.

Sub-Setting

#subset data
subset_ds_10 <-screen_time_ds[1:10, ] #The comma , separates rows from columns.

#convert it to a matrix
matrix_10 <- as.matrix(subset_ds_10)
print(matrix_10)

   Age Gender   Screen_Type    Day_Type  Avg_Screen_Time Sample_Size
1  "5" "Male"   "Educational"  "Weekday" "0.44"          "500"      
2  "5" "Male"   "Recreational" "Weekday" "1.11"          "500"      
3  "5" "Male"   "Total"        "Weekday" "1.55"          "500"      
4  "5" "Male"   "Educational"  "Weekend" "0.50"          "500"      
5  "5" "Male"   "Recreational" "Weekend" "1.44"          "500"      
6  "5" "Male"   "Total"        "Weekend" "1.93"          "500"      
7  "5" "Female" "Educational"  "Weekday" "0.49"          "500"      
8  "5" "Female" "Recreational" "Weekday" "0.96"          "500"      
9  "5" "Female" "Total"        "Weekday" "1.45"          "500"      
10 "5" "Female" "Educational"  "Weekend" "0.50"          "500"      

str(matrix_10)

 chr [1:10, 1:6] "5" "5" "5" "5" "5" "5" "5" "5" "5" "5" "Male" "Male" "Male" ...
 - attr(*, "dimnames")=List of 2
  ..$ : chr [1:10] "1" "2" "3" "4" ...
  ..$ : chr [1:6] "Age" "Gender" "Screen_Type" "Day_Type" ...

First, I had to subset data for first 10 observations with all variables. 1:10 means select rows from 1 to 10. The comma , separates rows from columns. Space after the comma means give all the columns. And I am saving the subset data into a new object call ‘subset_ds_10’.

Then to convert it for a matrix, I simply used the as.matrix() function to convert the data frame into a matrix and saved it in a new object called ‘matrix_10’.

A matrix can only hold one data type. In my data frame, I have both numbers and characters (like Age and Gender), R will convert everything into character strings in the matrix. So, we can see that after using the str(matrix_10), it gives the summary as all are chr now.

Create a new Data Frame

# Create integer variable (Number of minutes the child spends on physical activity)
physical_activity_minutes <- c(60,30,20,120,40,20,100,15,10,90)

# Create ordinal variable (Parental monitoring or restrictions over the child’s screen usage)
parental_control_level <- c("Low", "High", "High", "None", "Moderate", "High", "None", "High", "High", "Low") 

str(parental_control_level)

 chr [1:10] "Low" "High" "High" "None" "Moderate" "High" "None" "High" "High" ...

class(parental_control_level)

[1] "character"

# Convert parental_control_level to ordered factor
parental_control_level <- factor(parental_control_level, 
                                 levels = c("None", "Low", "Moderate", "High"),
                                 ordered = TRUE)
str(parental_control_level)

 Ord.factor w/ 4 levels "None"<"Low"<"Moderate"<..: 2 4 4 1 3 4 1 4 4 2

levels(parental_control_level)

[1] "None"     "Low"      "Moderate" "High"    

class(parental_control_level)

[1] "ordered" "factor" 

# Combine into data frame
child_df <- data.frame(Physical_Activity_Minutes = physical_activity_minutes, Parental_Control_Level = parental_control_level)

str(child_df)

'data.frame':   10 obs. of  2 variables:
 $ Physical_Activity_Minutes: num  60 30 20 120 40 20 100 15 10 90
 $ Parental_Control_Level   : Ord.factor w/ 4 levels "None"<"Low"<"Moderate"<..: 2 4 4 1 3 4 1 4 4 2

# Create numeric variable (Sleep Duration (hours))
sleep_duration_hours <- c(9.2,8.7, 7.5, 9.0, 6.8, 8.3, 7.0, 9.5, 6.5, 8.0)

#use cbind() to add this vector to your data frame.
child_df <- cbind(child_df, sleep_duration_hours)

#get dimensions 
dim(child_df)

[1] 10  3

str(child_df)

'data.frame':   10 obs. of  3 variables:
 $ Physical_Activity_Minutes: num  60 30 20 120 40 20 100 15 10 90
 $ Parental_Control_Level   : Ord.factor w/ 4 levels "None"<"Low"<"Moderate"<..: 2 4 4 1 3 4 1 4 4 2
 $ sleep_duration_hours     : num  9.2 8.7 7.5 9 6.8 8.3 7 9.5 6.5 8

Even though it is independent from the previous data set, I made one integer variable and ordinal variable which goes with my previous data set. For each,it has 10 observations. ‘parental_control_level’ class is character. As per instructed, I converted ordinal variable in to a ordered factor variable. Then I got the class as “ordered” “factor”. And using the data.frame (), I finally made the data frame using integer variable and ordered factor variable. Again created a numeric variable. Using cbind() function, I added the new numeric vector variable with the previous data frame. Using dim() function, it confirms that in the end there are 10 observations with 3 variables.

Presentation

Presentation link: https://rmit-arc.instructuremedia.com/embed/f20ce682-1340-4aba-b1a5-3ff3139a24d1

References

Stack overflow (2025), Stack overflow website, accessed 09 April 2025. https://stackoverflow.com/questions/21125222/determine-the-data-types-of-a-data-frames-columns

W3Schools (2025) R Tutorial, w3schools website, accessed 29 March 2025. https://www.w3schools.com/r/r_data_frames.asp

AKSHAY KUMAR (2025) Average Daily Screen Time for Children, Kaggle website, accessed 29 March 2025. https://www.kaggle.com/datasets/ak0212/average-daily-screen-time-for-children

DataCamp (2024) Subsetting in R Tutorial, datacamp website, 10 April 2025. https://www.datacamp.com/tutorial/subsets-in-r