Reproducible Research: Peer Assessment 1

Loading and preprocessing the data

## 
## Attaching package: 'zip'

## The following object is masked from 'package:utils':
## 
##     zip

What is mean total number of steps taken per day?

# 2. Histogram of the total number of steps taken each day
stepsPerDay <- tapply(X = df$step, INDEX = df$date, FUN = sum, na.rm = TRUE)
par(mfrow = c(1,1))
hist(stepsPerDay, breaks = 30, col = "red", xlab = "Steps per day", main = "Histogram of steps per day")

What is the average daily activity pattern?

# 3. Mean and median number of steps taken each day
meanStepsPerDay <- round(mean(stepsPerDay), digits = 2)
medianStepsPerDay <- round(median(stepsPerDay), digits = 2)

# 4. Time series plot of the average number of steps taken
par(mfrow = c(1,1))
averageStepsPerInterval <- tapply(X = df$step, INDEX = df$interval, FUN = mean, na.rm = T)
plot(averageStepsPerInterval, type = "l", col="magenta", main = "Average number of steps taken (ignoring missing values)", xlab = "Interval", ylab = "Average steps per interval")

# 5. The 5-minute interval that, on average, contains the maximum number of steps
intervalWithMaxValue <- which.max(averageStepsPerInterval)
intervalMaxValue <- round(max(averageStepsPerInterval), digits = 2)
maxIntervalName <- df$interval[intervalWithMaxValue[1]]

The mean number of steps per day is 9354.23.

The median number of steps per day is 10395.

The interval that on average across all the days in the dataset, contains the maximum number of steps is the 104-th that corresponds to interval 835 with a value of 206.17 steps per day.

Imputing missing values

# 6. Code to describe and show a strategy for imputing missing data: the mean for that 5-minute interval
numberOfNA <- sum(is.na(df$steps))

There are 2304 NA (missing values) in the steps column of the dataset.

The chosen strategy to impute a missing value is to use the mean of the corresponding 5-minute interval.

# Create a new dataset that is equal to the original dataset but with the missing data filled in.
imputeddf <- df
i <- -1
imputeddf$steps <- sapply(df$steps, FUN = function(x) {
    i <<- i + 1
    if(is.na(x)) averageStepsPerInterval[[i %% 288 +1]]
    else x
}
)

imputedstepsPerDay <- tapply(X = imputeddf$step, INDEX = imputeddf$date, FUN = sum, na.rm = TRUE)
# 7. Histogram of the total number of steps taken each day after missing values are imputed. 
hist(imputedstepsPerDay, breaks = 30, col = "yellow", main = "Histogram of mean of steps per day (with imputed missing values)", xlab = "Steps")

imputedMeanStepsPerDay   <- round(mean(imputedstepsPerDay), digits = 2)
imputedMedianStepsPerDay <- round(median(imputedstepsPerDay), digits = 2)

The mean and median of the total number of steps taken per day after imputing the missing value increase with respect to the corresponding values calculated ignoring missing values.

# Add spectacular-looking table to impress reviewer.

Tables	mean	median
Ignoring missing values	9354.23	10395
Imputing missing values	10766.19	10766.19

Are there differences in activity patterns between weekdays and weekends?

# Creating a new factor variable in the dataset with two levels – “weekday” and “weekend” indicating whether a given date is a weekday or weekend day.
findDayType <- function(date) {
    if(weekdays(date) == "sábado" | weekdays(date) == "domingo")
        "weekend"
    else
        "weekday"
} 
imputeddf$daytype = sapply(imputeddf$date, FUN = findDayType)

# 8. Panel plot comparing the average number of steps taken per 5-minute interval across weekdays and weekends

par(mfrow = c(2,1), mar = c(2,4,2,1))
weekdaydf <- imputeddf[imputeddf$daytype == "weekday",]
averageStepsPerIntervalWD <- tapply(X = weekdaydf$step, weekdaydf$interval, mean)
plot(averageStepsPerIntervalWD, type = "l", col = "blue", main = "Average number of steps (weekdays)",
     xlab = "Interval", ylab = "Number of steps")

weekenddf <- imputeddf[imputeddf$daytype == "weekend",]
averageStepsPerIntervalWE <- tapply(X = weekenddf$step, INDEX = weekenddf$interval, FUN = mean)
plot(averageStepsPerIntervalWE, type = "l", col = "green", main = "Average number of steps (weekends)",
     xlab = "Interval", ylab = "Number of steps")

It is easy to see that during the weekends the mean number of steps per day is higher than on weekdays.