RepData Peer Assessment Assignment

Loading and preprocessing the data

library(readr)
activity <- read_csv("C:/Users/FACPAN/Desktop/RepData_PeerAssessment1/activity/activity.csv")

## Parsed with column specification:
## cols(
##   steps = col_double(),
##   date = col_date(format = ""),
##   interval = col_double()
## )

View(activity)
head(activity)

## # A tibble: 6 x 3
##   steps date       interval
##   <dbl> <date>        <dbl>
## 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?

library(ggplot2)

## Registered S3 methods overwritten by 'ggplot2':
##   method         from 
##   [.quosures     rlang
##   c.quosures     rlang
##   print.quosures rlang

total.steps <- tapply(activity$steps, activity$date, FUN=sum, na.rm=TRUE)
qplot(total.steps, binwidth=1000, xlab="total number of steps taken each day")

mean(total.steps, na.rm=TRUE)

## [1] 9354.23

median(total.steps, na.rm=TRUE)

## [1] 10395

What is the average daily activity pattern?

library(ggplot2)
averages <- aggregate(x=list(steps=activity$steps), by=list(interval=activity$interval),
                      FUN=mean, na.rm=TRUE)
ggplot(data=averages, aes(x=interval, y=steps)) +
    geom_line() +
    xlab("5-minute interval") +
    ylab("average number of steps taken")

on the Average

averages[which.max(averages$steps),]

##     interval    steps
## 104      835 206.1698

Imputing missing values

missing <- is.na(activity$steps)

How many missing steps

table(missing)

## missing
## FALSE  TRUE 
## 15264  2304

The total count of NA in the dataset is 2304.

All of the missing values are filled in with mean value for that 5-minute

interval.

# Replace each missing value with the mean value of its 5-minute interval
fill.value <- function(steps, interval) {
    filled <- NA
    if (!is.na(steps))
        filled <- c(steps)
    else
        filled <- (averages[averages$interval==interval, "steps"])
    return(filled)
}
filled.activity <- activity
filled.activity$steps <- mapply(fill.value, filled.activity$steps, filled.activity$interval)

Now, using the filled data set, let’s make a histogram of the total number of steps taken each day and calculate the mean and median total number of steps.

total.steps <- tapply(filled.activity$steps, filled.activity$date, FUN=sum)
qplot(total.steps, binwidth=1000, xlab="total number of steps taken each day")

mean(total.steps)

## [1] 10766.19

median(total.steps)

## [1] 10766.19

Mean and median values are higher after imputing missing data. The reason is

that in the original data, there are some days with steps values NA for any interval. The total number of steps taken in such days are set to 0s by default. However, after replacing missing steps values with the mean steps of associated interval value, these 0 values are removed from the histogram of total number of steps taken each day.

Are there differences in activity patterns between weekdays and weekends?

First, let’s find the day of the week for each measurement in the dataset. In this part, we use the dataset with the filled-in values.

weekday.or.weekend <- function(date) {
    day <- weekdays(date)
    if (day %in% c("Monday", "Tuesday", "Wednesday", "Thursday", "Friday"))
        return("weekday")
    else if (day %in% c("Saturday", "Sunday"))
        return("weekend")
    else
        stop("invalid date")
}
filled.activity$date <- as.Date(filled.activity$date)
filled.activity$day <- sapply(filled.activity$date, FUN=weekday.or.weekend)

Now, let’s make a panel plot containing plots of average number of steps taken on weekdays and weekends.

Now, let’s make a panel plot containing plots of average number of steps taken on weekdays and weekends.

averages <- aggregate(steps ~ interval + day, filled.activity, mean)
ggplot(averages, aes(interval, steps)) + geom_line() + facet_grid(day ~ .) +
    xlab("5-minute interval") + ylab("Number of steps")