Activity monitoring data analysis

This is an R Markdown document for my solution to the assignment 1 of course Reproduciable Research on Coursera

Loading and preprocessing the data

Let’s take a first look of the dataset

data_path = "/Users/chenhao/GitProject/RepData_PeerAssessment1/"
activity_data = read.csv(paste(data_path,"activity.csv", sep=""), 
                         stringsAsFactors = F)
# format the data and transform NA to 0
activity_data$date <- as.Date(activity_data$date, "%Y-%m-%d")
#activity_data$steps[is.na(activity_data$steps)] = 0
head(activity_data)

##   steps       date interval
## 1    NA 2012-10-01        0
## 2    NA 2012-10-01        5
## 3    NA 2012-10-01       10
## 4    NA 2012-10-01       15
## 5    NA 2012-10-01       20
## 6    NA 2012-10-01       25

What is mean total number of steps taken per day?

split the data by date and do analysis of steps on date

# organize the date by days
sum_steps_per_day <- aggregate(steps ~ date, data = activity_data, sum)
mean_steps = as.integer(mean(sum_steps_per_day$step, na.rm = T))
median_steps = as.integer(median(sum_steps_per_day$step, na.rm = T))
# plot the histogram of the total number of steps taken each day
hist(sum_steps_per_day$steps, 
     main="Histogram of The Total Number of Steps Taken Each Day",
     xlab = "total number of steps taken per day"
     ) 

The mean total number of steps taken per day was 10766 , and The median total number of steps taken per day was 10765.

What is the average daily activity pattern?

split the data by interval and do analysis of steps on interval

# organize the data by daily time series
daily_steps <- aggregate(steps ~ interval, data = activity_data, mean)
plot(daily_steps$interval, 
     daily_steps$steps,
     type = "l",
     main = "Average Number of Steps Taken Each 5 Minutes Interval",
     xlab = "time interval",
     ylab = "average steps taken")

abline(v = daily_steps[daily_steps$steps == max(daily_steps$steps), 1], 
       col = "red")
text(x = daily_steps[daily_steps$steps == max(daily_steps$steps),1],
     y = max(daily_steps$steps),
     labels= max(daily_steps$steps),
     cex= 0.7, offset = 10)

maximun_activity_pattern  = daily_steps[daily_steps$steps == max(daily_steps$steps),1]

the 835th 5-minute interval contains the maximum number of steps on average across all the days in the dataset.

Imputing missing values

# check the number of NA values
NA_num = sum(is.na(activity_data$steps))

# fill the missing values with mean steps of that interval
new_activity_data <- activity_data
for (i in 1:length(new_activity_data$steps)) {
       if (is.na(new_activity_data$steps[i])) {
               mean_activity <- daily_steps[daily_steps$interval == new_activity_data$interval[i], 2]
               new_activity_data$steps[i] <- mean_activity }        
}

# Make a histogram of the total number of steps taken each day
new_sum_steps_per_day <- aggregate(steps ~ date, data = new_activity_data, sum)
new_mean_steps = as.integer(mean(new_sum_steps_per_day$step, na.rm = T))
new_median_steps = as.integer(median(new_sum_steps_per_day$step, na.rm = T))
# plot the histogram of the total number of steps taken each day
hist(new_sum_steps_per_day$steps, 
     main="Histogram of The Total Number of Steps Taken Each Day with filling of missing values",
     xlab = "total number of steps taken per day"
     ) 

After repalcing 2304 missing values with mean activity of that interval, The mean total number of steps taken per day was 10766 , and The median total number of steps taken per day was 10766. So we find that the mean value remain the same while the median increased a little.

Are there differences in activity patterns between weekdays and weekends?

Let’s split the data by week_label and time interval, then calculate the mean steps for each combine of weeklabel and time interval, using lattice to give a plot to visually show the answer.

weekdays = c("Monday", "Tuesday", "Wednesday", "Thursday", "Friday")
weekends = c("Saturday", "Sunday")

# add week labels to the dataset
week_label = vector()
new_activity_data$date <- weekdays(new_activity_data$date)
for (i in 1:length(new_activity_data$date)){
        if (new_activity_data$date[i] %in% weekdays){
                week_label = c(week_label, "weekday")  }
        else{
                week_label = c(week_label, "weekend")  }
}
new_activity_data <- transform( new_activity_data, week_labels = week_label)

# split the data by week label and interval then calcualte the mean steps, and plot
mean_steps_per_interval <- aggregate(steps ~ interval + week_labels, 
                                     data = new_activity_data, mean)
library(lattice)
xyplot(steps ~ interval|week_labels, 
       data = mean_steps_per_interval,
       type = "l",
       xlab = "Interval",
       ylab = "Number of steps",
       layout=c(1,2))

We can see that people get more activity on weekends, aha, due to more avaliable time.