Course Project 1 Reproducible Research

• 📚 Specialization: Data Science: Foundations using R Specialization
• 📖 Course: Reproducible Research
  • 🧑‍🏫 Instructor: Roger D Peng
• 📆 Week 2
  • 🚦 Start: Thursday, 23 June 2022
  • 🏁 Finish: Friday, 24 June 2022
• 📦 Github Repository: Static Document

---

Synopsis

Course Project 1 aims to answer questions and accomplish tasks to create a Reproducible Research. The dataset used for this assignment is the repdata_data_activity.zip, which has data about personal movement (steps):

• 279 kBytes (calculate using pryr package)
• 3 variables
• 17.568 observations.

The dataset required manipulation to clean the NA observation or define a strategy to deal with it. I have used an approach that considers the different behavior during the weekdays in its interval. The result of this imputing was an increase in the mean and median.

In the process of Exploratory Data Analysis, it was possible to identify that the tracked person has distinct behavior on Mondays, Tuesdays, and Wednesdays when they wake up early compared to Thursday and Friday, days they used to wake up at 8 AM. It was also possible to identify that on Weekends, they do not wake up early nor have intense physical exercises.

Finally, this tracked person used to make almost 11 thousand steps on average every day.

---

Course Project 1

Course Project 1 has five main tasks, and each task has its own subtasks:

• Task 1 – Loading and pre-processing the data;
• Task 2 – What is mean total number of steps taken per day?;
• Task 3 – What is the average daily activity pattern?;
• Task 4 – Imputing missing values;
• Task 5 – Are there differences in activity patterns between weekdays and weekends?.

Packages Requirements

It is necessary to install the following packages to reproduce this Course Project 2.

# Loading libraries
library(ggplot2)
library(tidyverse)
library(magrittr)
library(kableExtra)
library(stringr)
library(lubridate)

Also, those following scripts will use the standard packages from R.

1. base
2. graphic
3. utils
4. grDevices

Task 1

The Loading and pre-processing the data is divided into 2 subtasks:

• Subtask 1.1. – Load the data, and;
• Subtask 1.2. – Pre-processing the data.

I have moved the activity.zipfile to the data folder to make the repository root tidier.

Subtask 1.1.

Load the data

# Unzipping the compressed file
utils::unzip(zipfile = "./data/activity.zip",
             overwrite = TRUE,
             exdir = "./data/unzipped", # Extracting the zipped files into the unzipped folder.
             list = FALSE)

# Loading the .csv file
raw_dataset <- utils::read.csv(file = "./data/unzipped/activity.csv",
                               header = TRUE)

Following the principles of Exploratory Data Analysis, let’s dig into this dataset to analyze if it is necessary to make any adjustments. The first step is to know how many observations and variables this dataset has.

# Printing the dimensions.
base::dim(raw_dataset) # Rows Columns

## [1] 17568     3

The database has 17.568 observations and 3 variables. It is aligned with the Course Project 1 instructions.

Now, I will check those variables because, according to the assignment instruction, those variables should be steps, date, and interval.

# Print the column names.
base::colnames(raw_dataset)

## [1] "steps"    "date"     "interval"

The dataset has the expected variables and observations.

Subtask 1.2.

Pre-processing the data

Initially, I will display the dataset summary (summary()) and then the structure (str()). It is necessary to have a big picture of the data in each variable because some data could be missing or wrongly classified/imported.

# Printing the summary of each variables
summary(raw_dataset)

##      steps            date              interval     
##  Min.   :  0.00   Length:17568       Min.   :   0.0  
##  1st Qu.:  0.00   Class :character   1st Qu.: 588.8  
##  Median :  0.00   Mode  :character   Median :1177.5  
##  Mean   : 37.38                      Mean   :1177.5  
##  3rd Qu.: 12.00                      3rd Qu.:1766.2  
##  Max.   :806.00                      Max.   :2355.0  
##  NA's   :2304

# Printing the dataset structure.
str(raw_dataset)

## 'data.frame':    17568 obs. of  3 variables:
##  $ steps   : int  NA NA NA NA NA NA NA NA NA NA ...
##  $ date    : chr  "2012-10-01" "2012-10-01" "2012-10-01" "2012-10-01" ...
##  $ interval: int  0 5 10 15 20 25 30 35 40 45 ...

As warned by the Course Project 1 instructions, the steps variable has 2304 NA and is classified as an integer, the date variable is wrongly imported as “character”, and the interval has no NA and is classified as an integer.

It is necessary two steps:

1. Removing any observation with NA, and;
2. Converting the date column into a date class.

# STEP 1 - Removing observations with NA.
aux_dataset <- base::subset(x = raw_dataset, !base::is.na(raw_dataset$steps))

# STEP 2 - Converting date column to Date class
tidy_dataset <- aux_dataset %>%
    dplyr::mutate(date = base::as.Date(date))

# Tidy dataset dimensions
base::dim(tidy_dataset)

## [1] 15264     3

The tidy dataset has 15.264 observations with the same 3 variables.

# Printing the summary of each variables
summary(tidy_dataset)

##      steps             date               interval     
##  Min.   :  0.00   Min.   :2012-10-02   Min.   :   0.0  
##  1st Qu.:  0.00   1st Qu.:2012-10-16   1st Qu.: 588.8  
##  Median :  0.00   Median :2012-10-29   Median :1177.5  
##  Mean   : 37.38   Mean   :2012-10-30   Mean   :1177.5  
##  3rd Qu.: 12.00   3rd Qu.:2012-11-16   3rd Qu.:1766.2  
##  Max.   :806.00   Max.   :2012-11-29   Max.   :2355.0

# Printing the dataset structure.
str(tidy_dataset)

## 'data.frame':    15264 obs. of  3 variables:
##  $ steps   : int  0 0 0 0 0 0 0 0 0 0 ...
##  $ date    : Date, format: "2012-10-02" "2012-10-02" ...
##  $ interval: int  0 5 10 15 20 25 30 35 40 45 ...

The interval variable is recorded weirdly because each observation has a length of 4. The first two characters represent the hour and the last two the minutes. I will convert this column into a period class from lubridate package and then to hms.

Remember that the original interval has notation based on: “HHMM”. Where “HH” is related to the hour and “MM” to the minutes.

Example

Interval = 2350

It means the gadget has measured the steps between 23:50 and 23:55.

# Converting interval into Period Class.
tidy_dataset <- tidy_dataset %>%
    mutate(interval = str_pad(interval, 4, pad = "0")) %>%
    mutate(interval = paste0(substr(interval, 1, 2),":",substr(interval, 3, 4),":00")) %>%
    mutate(interval = hms::hms(lubridate::hms(interval))) # The S3 Period time is easier to work on ggplot2.

Finally, the tidy dataset has no NA, the date column is correctly classified as a Date class and interval variable is a “time” variable.

# Presenting the tidy dataset first rows.
head(tidy_dataset) %>%
    kableExtra::kbl() %>%
    kableExtra::kable_styling()

	steps	date	interval
289	0	2012-10-02	00:00:00
290	0	2012-10-02	00:05:00
291	0	2012-10-02	00:10:00
292	0	2012-10-02	00:15:00
293	0	2012-10-02	00:20:00
294	0	2012-10-02	00:25:00

Task 2

Task 2 entails about answer the question:

What is mean total number of steps taken per day?

In addition, there are also 3 subtasks:

• Subtask 2.1. – Calculate the total number of steps taken per day;
• Subtask 2.2. – Make a histogram of the total number of steps taken each day, and;
• Subtask 2.3. – Calculate the mean and median of daily steps.

The tidy_dataset has lines corresponding to a sample of 5 minutes of the day. If everything is correct, each day must have 288 observations.

# Checking if each day has 288 observations.
tidy_dataset %>%
    dplyr::group_by(date) %>%
    dplyr::summarise(obs_day = dplyr::n()) %>%  # Count the number of observations per day.
    dplyr::select(obs_day) %>%
    unique() %>%
    as.numeric()

## [1] 288

As you can see, all days have the same quantity of observations (288), allowing me to make further analyses using all observations of tidy_dataset.

Subtask 2.1.

Calculate the total number of steps taken per day

Now, it is necessary to aggregate the rows of the same day to calculate a dataset based on days.

# Calculating the number of steps in a day.
tidy_dataset_day <- tidy_dataset %>%
    dplyr::group_by(date) %>%
    dplyr::summarise(steps = sum(steps))

# ALTERNATIVE - Using the aggregate function
# tidy_dataset_day <- stats::aggregate(steps ~ date, data = tidy_dataset, FUN = sum)

According to the tidy_dataset there are 53 days in this dataset, so the tidy_dataset_day must have 53 rows.

# Number of rows
base::nrow(tidy_dataset_day)

## [1] 53

The tidy_dataset_day has 53 observations and 2 columns (date and steps). The scatter plot below will show the distribution of daily steps between 2012-10-02 and 2012-11-29.

# Plotting a scatter plot using ggplot2.
ggplot2::ggplot(data = tidy_dataset_day,
                aes(x = date,
                    y = steps)) + 
    
    # Scatter plot
    geom_point() +
    
    # Adding Loess Regression 
    geom_smooth(formula = y ~ x,
                orientation = "x",
                method = 'loess',
                lwd = 0.75,
                aes(color = "Loess")) + 
    
    # Adding a horizontal line to show the Average
    geom_hline(yintercept = mean(tidy_dataset_day$steps),
               linetype = "dashed",
               color = "red",
               lwd = 1,
               aes(color = "Average")) +
    
    # Adding manually the legend.
    scale_color_manual(name = '',
                       breaks = c('Loess', 'Average'),
                       values = c('Loess' = 'blue', 'Average' = 'red')) + 
    
    # Defining the Plot title, x-axis label and y-axis label.
    labs(title = "Time Series of Steps") + 
    xlab(label = "Day") +
    ylab(label = "Daily Steps") +
    
    # Setting legend and title positions.
    theme(legend.position = "bottom",
          plot.title = element_text(hjust = 0.5))

Subtask 2.2.

Make a histogram of the total number of steps taken each day

Based on tidy_dataset_day, the histogram below presents the daily steps habits.

# Plotting a ggplot2 histogram.
ggplot2::ggplot(data = tidy_dataset_day,
                aes(x = steps)) + 
    
    # Creating a histogram with 10 bins.
    geom_histogram(bins = 10) + 
    
    # Adding the Average
    geom_vline(xintercept = mean(tidy_dataset_day$steps),
               linetype = "dashed",
               color = "red",
               lwd = 1.0,
               aes(color = "Average")) + 
    
    # Defining the Plot title, x-axis label and y-axis label.
    labs(title = "Daily Steps Histogram",
         subtitle = "Removing NA") + 
    xlab(label = "Steps") +
    ylab(label = "Frequency") +
    
    # Setting legend and title positions.
    theme(legend.position = "right",
          plot.title = element_text(hjust = 0.5),
          plot.subtitle = element_text(hjust = 0.5))

According to the above histogram, the most common daily steps varies between 10.000 and 15.000 thousand steps.

Subtask 2.3.

Calculate the mean and median of daily steps.

The summary show the mean and median of daily steps.

# Summary of steps
summary(tidy_dataset_day) %>%
    kableExtra::kbl() %>%
    kableExtra::kable_styling()

	date	steps
	Min. :2012-10-02	Min. : 41
	1st Qu.:2012-10-16	1st Qu.: 8841
	Median :2012-10-29	Median :10765
	Mean :2012-10-30	Mean :10766
	3rd Qu.:2012-11-16	3rd Qu.:13294
	Max. :2012-11-29	Max. :21194

Finally, the daily steps mean and median:

• Mean: 10766
• Median: 10765

Task 3

Task 3 entails about answer the question:

What is the average daily activity pattern?

In addition, there are also two subtasks:

• Make a time series plot of the 5-minute interval (x-axis) and the average number of steps taken, averaged across all days (y-axis)
• Which 5-minute interval, on average across all the days in the dataset, contains the maximum number of steps?

The interval variable is responsible for tracking the daily activity pattern. For this reason, I will use it to create a new dataset of daily activity pattern.

# Calculating the number of steps in a day.
tidy_dataset_day_pattern <- tidy_dataset %>%
    dplyr::group_by(interval) %>%
    dplyr::summarise(steps = mean(steps))

# ALTERNATIVE - Using the aggregate function
# tidy_dataset_day_pattern <- base::aggregate(steps ~ interval, data = tidy_dataset, FUN = sum)

Let’s check how is the dataset after the average. I will display the last rows.

# Presenting the Daily Pattern last rows.
tail(tidy_dataset_day_pattern) %>%
    kableExtra::kbl() %>%
    kableExtra::kable_styling()

interval	steps
23:30:00	2.6037736
23:35:00	4.6981132
23:40:00	3.3018868
23:45:00	0.6415094
23:50:00	0.2264151
23:55:00	1.0754717

Subtask 3.1.

Make a time series plot of the 5-minute interval (x-axis) and the average number of steps taken, averaged across all days (y-axis)

The line plot below presents the average daily steps in each interval.

# Creating a plotting with line and points
ggplot(data = tidy_dataset_day_pattern) + 
    
    # Early morning period.
    geom_rect(aes(xmin = -Inf,
                  xmax = hms::hms(lubridate::hms("06:00:00")),
                  ymin= -Inf,
                  ymax = Inf), fill = "lightblue", alpha = 0.015) + 
    
    # Daylight period.
    geom_rect(aes(xmin = hms::hms(lubridate::hms("06:00:00")),
                  xmax = hms::hms(lubridate::hms("18:00:00")),
                  ymin= -Inf,
                  ymax = Inf), fill = "yellow", alpha = 0.015) +
    
    # Night time period.
    geom_rect(aes(xmin = hms::hms(lubridate::hms("18:00:00")),
                  xmax = Inf,
                  ymin= -Inf,
                  ymax = Inf), fill = "lightblue", alpha = 0.015) +   
    
    # Adding lines.
    geom_line(aes(x = interval, y = steps)) + 
    
    # Adding points.
    geom_point(aes(x = interval, y = steps)) +
    
    # Defining manually the breaks.
    scale_x_time(breaks = c(hms::hms(lubridate::hms("00:00:00")),
                                           hms::hms(lubridate::hms("06:00:00")),
                                           hms::hms(lubridate::hms("12:00:00")),
                                           hms::hms(lubridate::hms("18:00:00")),
                                           hms::hms(lubridate::hms("24:00:00")))) + 

    # Defining the Plot title, x-axis label and y-axis label.
    labs(title = "Time Serie of Daily Interval Steps Average") + 
    xlab(label = "Interval") +
    ylab(label = "Steps Average") +
    
    # Setting legend and title positions.
    theme(plot.title = element_text(hjust = 0.5))

The person tracked has an explicit behavior to wake up around 5 AM, between 7 AM and 10 AM has the peak of steps (maybe the commute to work), around mid-day another rise which could be the trajectory to lunch.

Subtask 3.2.

Which 5-minute interval, on average across all the days in the dataset, contains the maximum number of steps?

Let’s filter the tidy_dataset_day_pattern to obtain the highest steps average and find when it happens.

# Printing the maximum steps average and the interval what happens.
tidy_dataset_day_pattern %>%
    filter(steps == max(steps)) %>%
    kableExtra::kbl() %>%
    kableExtra::kable_styling()

interval	steps
08:35:00	206.1698

The highest steps average was 206.17, and it occurred at 08:35 AM.

Task 4

The Imputing missing values task implies replacing the NA value with another value instead of removing the entire observation. This tasks has four subtasks:

• Calculate and report the total number of missing values in the dataset (i.e. the total number of rows with NAs);
• Devise a strategy for filling in all of the missing values in the dataset. The strategy does not need to be sophisticated. For example, you could use the mean/median for that day, or the mean for that 5-minute interval, etc;
• Create a new dataset that is equal to the original dataset but with the missing data filled in, and;
• Make a histogram of the total number of steps taken each day and Calculate and report the mean and median total number of steps taken per day. Do these values differ from the estimates from the first part of the assignment? What is the impact of imputing missing data on the estimates of the total daily number of steps?

Subtask 4.1.

Calculate and report the total number of missing values in the dataset (i.e. the total number of rows with NAs)

It is already answered in Subtask 1.2., however, I will put the summary() one more time.

# Summary will display an overview of all variables.
summary(raw_dataset)

##      steps            date              interval     
##  Min.   :  0.00   Length:17568       Min.   :   0.0  
##  1st Qu.:  0.00   Class :character   1st Qu.: 588.8  
##  Median :  0.00   Mode  :character   Median :1177.5  
##  Mean   : 37.38                      Mean   :1177.5  
##  3rd Qu.: 12.00                      3rd Qu.:1766.2  
##  Max.   :806.00                      Max.   :2355.0  
##  NA's   :2304

# Counting the NA number in steps column.
sum(is.na(raw_dataset$steps))

## [1] 2304

The steps variables has 2304 NA observations.

Subtask 4.2.

Devise a strategy for filling in all of the missing values in the dataset. The strategy does not need to be sophisticated. For example, you could use the mean/median for that day, or the mean for that 5-minute interval, etc.

Each day of the week has its characteristic for anyone, e.g., sometimes on Mondays we have to go to the gym, on Tuesdays we usually have Languages Courses, and so on. Based on it, I have defined my strategy as calculating an average for each weekday divided into intervals.

First, I have to perform almost the same process as Subtask 1.2., except not removing the NA observations.

# Creating the Tidy dataset with NA will be updated later, replacing NA with other values.
tidy_dataset_2 <- raw_dataset %>%
    
    # Converting date as Date class.
    dplyr::mutate(date = base::as.Date(date)) %>%
    
    # Adding a new column of Weekdays as a factor.
    mutate(weekday = factor(weekdays(date),
                            levels = c("Sunday", "Monday", "Tuesday", "Wednesday", "Thursday", "Friday", "Saturday"))) %>%
    
    # Manipulating the interval column to convert it as S3 Period.
    mutate(interval = str_pad(interval, 4, pad = "0")) %>%
    mutate(interval = paste0(substr(interval, 1, 2),":",substr(interval, 3, 4),":00")) %>%
    mutate(interval = hms::hms(lubridate::hms(interval)))

Now, I will calculate the average steps for each interval divided into weekdays.

# Auxiliary Dataset to store the average step for each weekday and interval.
avg_weekday <- tidy_dataset_2 %>%
    group_by(weekday, interval) %>%
    summarise(average = mean(steps,
                             na.rm = TRUE))

Based on avg_weekday, it is possible to plot a line plot showing the differences between each weekday.

# Data manipulation to create new features.
avg_weekday %>%
    
    # Calculating the average of each interval.
    group_by(interval,weekday) %>%
    summarise(steps = mean(average)) %>%
    
    # Creating the ggplot2 graphic.
    ggplot(aes(x = interval,
               y = steps)) + 
    
    # Adding lines.
    geom_line() + 
    
    # Adding points.
    geom_point() + 
    
    # Defining manually the breaks.
    scale_x_time(breaks = c(hms::hms(lubridate::hms("00:00:00")),
                            hms::hms(lubridate::hms("06:00:00")),
                            hms::hms(lubridate::hms("12:00:00")),
                            hms::hms(lubridate::hms("18:00:00")),
                            hms::hms(lubridate::hms("24:00:00")))) + 
    
    # Defining the Plot title, x-axis label and y-axis label.
    labs(title = "Time Series of Daily Interval Steps Average") + 
    xlab(label = "Interval [hour]") +
    ylab(label = "Steps Average") +
    
    # Setting legend and title positions.
    theme(plot.title = element_text(hjust = 0.5)) + 
    
    # Creating facets to show weekdays differences.
    facet_grid(rows = vars(weekday))

The tracked person has recorded activities since 6 AM on Monday, Tuesday, and Wednesday. On all other days, they start to record actions from 8 AM.

Just for curiosity, I want to know from those NA observations, how is the distribution between weekdays?

# Counting days with no observations.
tidy_dataset_2 %>%
    group_by(weekday) %>%
    summarise(days = sum(is.na(steps))/288)

## # A tibble: 7 x 2
##   weekday    days
##   <fct>     <dbl>
## 1 Sunday        1
## 2 Monday        2
## 3 Tuesday       0
## 4 Wednesday     1
## 5 Thursday      1
## 6 Friday        2
## 7 Saturday      1

As you can see, there is no concentration of NA on a single weekday.

Subtask 4.3.

Create a new dataset that is equal to the original dataset but with the missing data filled in.

Following the strategy defined in Subtask 4.2., I need to merge() two datasets: tidy_dataset_2 and avg_weekday, then I may fill in NA using the coalesce() function.

# Merging datasets using two keys.
merge(x = tidy_dataset_2,
      y = avg_weekday,
      by = c("interval", "weekday")) %>%
    
    # Arranging the dataset to be on crescent time.
    arrange(date) %>%
    
    # Fill in NA by average of each weekday and interval.
    mutate(steps = coalesce(steps, average)) -> tidy_dataset_2

Checking if all NA is replaced with average.

# Checking NA
sum(is.na(tidy_dataset_2$steps))

## [1] 0

According to summary() there’s no NA in interval column.

Subtask 4.4.

Make a histogram of the total number of steps taken each day and Calculate and report the mean and median total number of steps taken per day. Do these values differ from the estimates from the first part of the assignment? What is the impact of imputing missing data on the estimates of the total daily number of steps?

The code below will display a histogram using the dataset with NA fill-in.

# Calculating the summation of steps in a day.
tidy_dataset_2 %>% group_by(date) %>%
    summarise(steps = sum(steps)) -> tidy_dataset_aggregate

# Plotting a ggplot2 histogram.
ggplot2::ggplot(data = tidy_dataset_aggregate,
                aes(x = steps)) + 
    
    # Creating a histogram with 10 bins.
    geom_histogram(bins = 10) + 
    
    # Adding the Average
    geom_vline(xintercept = mean(tidy_dataset_aggregate$steps),
               linetype = "dashed",
               color = "red",
               lwd = 1.0,
               aes(color = "Average")) + 
    
    # Defining the Plot title, x-axis label and y-axis label.
    labs(title = "Daily Steps Histogram",
         subtitle = "Fill-in NA") + 
    xlab(label = "Steps") +
    ylab(label = "Frequency") +
    
    # Setting legend and title positions.
    theme(legend.position = "right",
          plot.title = element_text(hjust = 0.5),
          plot.subtitle = element_text(hjust = 0.5))

Let’s check the new values of the mean and median after the fill-in process.

summary(tidy_dataset_aggregate)

##       date                steps      
##  Min.   :2012-10-01   Min.   :   41  
##  1st Qu.:2012-10-16   1st Qu.: 8918  
##  Median :2012-10-31   Median :11015  
##  Mean   :2012-10-31   Mean   :10821  
##  3rd Qu.:2012-11-15   3rd Qu.:12811  
##  Max.   :2012-11-30   Max.   :21194

The mean and median from fill-in NA and removed NA are different. The table below presents the difference between those results.

	Removing NA	Fill-in NA
mean	10766.19	10821.21
median	10765.00	11015.00

The fill-in NA impact was a mean and median slightly higher.

Task 5

Are there differences in activity patterns between weekdays and weekends?

Task 5 aims to find patterns in the dataset, and it is divided into two subtasks:

• Create a new factor variable in the dataset with two levels – “weekday” and “weekend” indicating whether a given date is a weekday or weekend day, and;
• Make a panel plot containing a time series plot (i.e. type = “l”) of the 5-minute interval (x-axis) and the average number of steps taken, averaged across all weekday days or weekend days (y-axis). See the README file in the GitHub repository to see an example of what this plot should look like using simulated data.

Subtask 5.1.

Create a new factor variable in the dataset with two levels – “weekday” and “weekend” indicating whether a given date is a weekday or weekend day.

Following the Course Project 1 instructions, I need to create a new column storing the info about Weekend or Weekday.

# Adding a new variable about weekend or weeday.
tidy_dataset_2 %>%
    mutate(day_type = case_when(
        weekday %in% c("Friday", "Saturday") ~ factor("Weekend", levels = c("Weekend", "Weekday")),
        !(weekday %in% c("Friday", "Saturday")) ~ factor("Weekday", levels = c("Weekend", "Weekday")))) -> tidy_dataset_2

Let’s see the new variable called day_type:

interval	weekday	steps	date	average	day_type
00:00:00	Monday	1.428571	2012-10-01	1.428571	Weekday
00:05:00	Monday	0.000000	2012-10-01	0.000000	Weekday
00:10:00	Monday	0.000000	2012-10-01	0.000000	Weekday
00:15:00	Monday	0.000000	2012-10-01	0.000000	Weekday
00:20:00	Monday	0.000000	2012-10-01	0.000000	Weekday
00:25:00	Monday	5.000000	2012-10-01	5.000000	Weekday

Finally, following the instruction, the new variable data type is a factor.

# Checking the new variable. It must be a factor.
str(tidy_dataset_2)

## 'data.frame':    17568 obs. of  6 variables:
##  $ interval: 'hms' num  00:00:00 00:05:00 00:10:00 00:15:00 ...
##   ..- attr(*, "units")= chr "secs"
##  $ weekday : Factor w/ 7 levels "Sunday","Monday",..: 2 2 2 2 2 2 2 2 2 2 ...
##  $ steps   : num  1.43 0 0 0 0 ...
##  $ date    : Date, format: "2012-10-01" "2012-10-01" ...
##  $ average : num  1.43 0 0 0 0 ...
##  $ day_type: Factor w/ 2 levels "Weekend","Weekday": 2 2 2 2 2 2 2 2 2 2 ...

Subtask 5.2.

Make a panel plot containing a time series plot (i.e. type = “l”) of the 5-minute interval (x-axis) and the average number of steps taken, averaged across all weekday days or weekend days (y-axis). See the README file in the GitHub repository to see an example of what this plot should look like using simulated data.

Creating a plot line with two panels: Weekend and Weekdays.

# Calculating the mean of all weekends days and weekdays.
tidy_dataset_2 %>% 
    group_by(interval, day_type) %>%
    summarise(steps = mean(steps)) %>%

    # Plotting a geom_line e geom_poiny.
ggplot(aes(x = interval, y = steps)) + 
    
        # Early morning period.
    geom_rect(aes(xmin = -Inf,
                  xmax = hms::hms(lubridate::hms("06:00:00")),
                  ymin= -Inf,
                  ymax = Inf), fill = "lightblue", alpha = 0.015) + 
    
    # Daylight period.
    geom_rect(aes(xmin = hms::hms(lubridate::hms("06:00:00")),
                  xmax = hms::hms(lubridate::hms("18:00:00")),
                  ymin= -Inf,
                  ymax = Inf), fill = "yellow", alpha = 0.015) +
    
    # Night time period.
    geom_rect(aes(xmin = hms::hms(lubridate::hms("18:00:00")),
                  xmax = Inf,
                  ymin= -Inf,
                  ymax = Inf), fill = "lightblue", alpha = 0.015) + 
    
    geom_point() + 
    
    geom_line() + 
    
    # Defining manually the breaks.
    scale_x_time(breaks = c(hms::hms(lubridate::hms("00:00:00")),
                                           hms::hms(lubridate::hms("06:00:00")),
                                           hms::hms(lubridate::hms("12:00:00")),
                                           hms::hms(lubridate::hms("18:00:00")),
                                           hms::hms(lubridate::hms("24:00:00")))) + 
    
    # Defining the Plot title, x-axis label and y-axis label.
    labs(title = "Time Serie of Daily Interval Steps Average",
         subtitle = "Divided into Weekend and Weekdays") + 
    xlab(label = "Interval") +
    ylab(label = "Steps Average") + 
    
    # Setting legend and title positions.
    theme(plot.title = element_text(hjust = 0.5),
          plot.subtitle = element_text(hjust = 0.5)) + 
    
    # Creating the facet grid by weekdays and weekend.
    facet_grid(rows = vars(day_type))

During the weekdays, the user has a different daily routine. For example, on Weekends, they wake up later and tend to be less active. However, we have already dug into the weekdays and found out the user is more active on Monday, Tuesday, and Wednesday, as presented in Subtask 4.2..