Loading and preprocessing the data

# Unzip the file.
unzip("./activity.zip")

# Load the data file into a data frame.
raw_data <- read.csv("./activity.csv")

# Take a look at raw_data.
head(raw_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?

# Calculate the total number of steps per day.
step_per_day <- aggregate(steps ~ date, raw_data, sum)

# Take a look at step_per_day.
head(step_per_day)
##         date steps
## 1 2012-10-02   126
## 2 2012-10-03 11352
## 3 2012-10-04 12116
## 4 2012-10-05 13294
## 5 2012-10-06 15420
## 6 2012-10-07 11015
# Plot a histogram.
hist(step_per_day$steps, main = "Histogram of the Total Number of Steps Per Day", xlab = "Steps Per Day")

# Calculate the mean.
round(mean(step_per_day$steps))
## [1] 10766
# Calculate the median.
median(step_per_day$steps)
## [1] 10765

The mean total number of steps taken per day is \(10,766\) and the median is \(10,765\).

What is the average daily activity pattern?

# Calculate the average number of steps per interval.
step_by_interval <- aggregate(steps ~ interval, raw_data, mean)

# Take a look at step_by_inverval.
head(step_by_interval)
##   interval     steps
## 1        0 1.7169811
## 2        5 0.3396226
## 3       10 0.1320755
## 4       15 0.1509434
## 5       20 0.0754717
## 6       25 2.0943396
# Make a time series plot of the 5-minute interval (x-axis) and the average number of steps taken (y-axis).
plot(step_by_interval$interval, step_by_interval$steps, xlab = "5-minute Interval Index", ylab = "Average Number of Steps", type = "l")

# Find the 5-minute interval that contains the maximum.
step_by_interval$interval[which.max(step_by_interval$steps)]
## [1] 835
# Find the maximum number of steps in that interval.
round(step_by_interval$steps[which.max(step_by_interval$steps)])
## [1] 206

The maximum number of steps is \(206\) in the \(835^{th}\) interval.

Imputing missing values

# Calculate and report the total number of missing values in the dataset.
na_rows <-is.na(raw_data$steps)
sum(na_rows)
## [1] 2304
# Fill in all of the missing values in the dataset using the the mean for that 5-minute interval, etc.
filled_data <- split(raw_data, raw_data$interval)

for (i in 1:288) {
  filled_data[[i]]$steps[is.na(filled_data[[i]]$steps)]=step_by_interval$steps[i]
}

library(plyr)
new_data <- ldply(filled_data)[,-1]

# Make a histogram of the total number of steps taken each day.
new_step_per_day <- aggregate(steps ~ date, new_data, sum)
hist(new_step_per_day$steps, main = "Histogram of the Total Number of Steps Per Day", xlab = "Steps Per Day")

# Calculate the mean.
round(mean(new_step_per_day$steps))
## [1] 10766
# Calculate the median.
median(new_step_per_day$steps)
## [1] 10766.19

The new mean total number of steps taken per day is \(10,766\) and the new median is \(10,766.19\).

The new mean and median do NOT differ from the estimates from the first part of the assignment. The imputed data does not create obvious impact.

Are there differences in activity patterns between weekdays and weekends?

# Create a function to check day in week.
checkDayInWeek <- function(date) {
    d <- weekdays(as.Date(date, '%Y-%m-%d'))
    if  (!(d == 'Saturday' || d == 'Sunday')) {
        return('Weekday') 
    } else {
        return('Weekend')
    }
}

# Create a new factor variable  with two levels – “weekday” and “weekend”.
new_data$day <- as.factor(sapply(new_data$date, checkDayInWeek))

# Take a look at new_data.
head(new_data)
##       steps       date interval     day
## 1  1.716981 2012-10-01        0 Weekday
## 2  0.000000 2012-10-02        0 Weekday
## 3  0.000000 2012-10-03        0 Weekday
## 4 47.000000 2012-10-04        0 Weekday
## 5  0.000000 2012-10-05        0 Weekday
## 6  0.000000 2012-10-06        0 Weekend
## Calculate the average number of steps per interval by day type.
steps_by_day_type <- aggregate(steps ~ interval+day, new_data, mean)

# Make a panel plot containing a time series plotof the 5-minute interval (x-axis) and the average number of steps taken, averaged across all weekday days or weekend days (y-axis). 
library(ggplot2)
panel_plot <- ggplot(steps_by_day_type, aes(interval, steps)) +
    geom_line(stat = "identity", aes(colour = day)) +
    theme_gray() +
    facet_grid(day ~ ., scales="fixed", space="fixed") +
    labs(x="5-minute Interval Indexl", y=expression("Average No. of Steps")) +
    ggtitle("No. of Steps Per Interval by Day Type")
print(panel_plot)