Skip to content

Latest commit

 

History

History
188 lines (149 loc) · 6.08 KB

PA1_template.md

File metadata and controls

188 lines (149 loc) · 6.08 KB

PA1_template.Rmd

Reproducible Research Course

Peer Assessment 1

Load and preprocess the data.

setwd("C:/Users/Agilex01/Documents/Data")
rawtable <- read.csv('activity.csv', header = TRUE, sep = ",")
by_date <- group_by(rawtable, date)
num_steps <- summarise(by_date, sum(steps, na.rm = TRUE))
colnames(num_steps) <- c("date", "steps")
num_steps$date <- as.Date(num_steps$date)

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

bin <- 2
p <- ggplot(num_steps, aes(date, steps))
p <- p + geom_histogram(stat = "identity", binwidth = bin, colour = "white")
p <- p + scale_x_date(breaks = seq(min(num_steps$date)-10,
                                   max(num_steps$date),
                                   bin),
                      labels = date_format("%m-%d-%Y"),
                      limits = c(as.Date("2012-10-01"),
                                 as.Date("2012-11-30")))
p <- p + theme_bw() + xlab(NULL) + theme(axis.text.x  = element_text(angle=45,
                                                                  hjust = 1,
                                                                  vjust = 1))
p <- p + labs(title = "Total Number of Steps Taken per Day Using Raw Data") 
p

Calculate and report the mean and median of the total number of steps taken each day.

cat("The mean of the total number of steps is ", mean(num_steps$steps, na.rm = TRUE))
## The mean of the total number of steps is  9354.23

cat("The median of the total number of steps is ", median(num_steps$steps, na.rm = TRUE))
## The median of the total number of steps is  10395

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).

by_interval <- group_by(rawtable, interval)
mean_interval <- summarise(by_interval, mean(steps, na.rm = TRUE))
colnames(mean_interval) <- c("interval", "meansteps")
plot(mean_interval$interval, mean_interval$meansteps, xaxp  = c(0, 2400, 100), 
     xlab = "5 Minute Interval", ylab = "Average Number of Steps", 
     main = "Average Number of Steps from 10-1-12 to 11-30-12", type = "l")

sort_mean_interval <- mean_interval[order(-mean_interval$meansteps),]
max_interval <- as.numeric(sort_mean_interval[1,1])
cat("The 5 minute interval that has the maximum number of steps is", max_interval)
## The 5 minute interval that has the maximum number of steps is 835

num_nas <- colSums(is.na(rawtable))
cat("The number of missing values in the dataset is", num_nas[1])
## The number of missing values in the dataset is 2304

Create a new imputed data set with the missing values filled in. The strategy to fill in the missing data used the mean of each interval.

imput_table <- rawtable
for(i in 1:dim(imput_table)[1])
        if (is.na(imput_table[i,1]))
                  imput_table[i,1] = mean_interval[which(mean_interval$interval == imput_table[i,3]),2]

Calculate the total number of steps taken each day.

imput_by_date <- group_by(imput_table, date)
imput_num_steps <- summarise(imput_by_date, sum(steps))

Make a histogram of the total number of steps taken each day using the imputed data.

colnames(imput_num_steps) <- c("date", "steps")
imput_num_steps$date <- as.Date(imput_num_steps$date)
bin <- 2
imput_plot <- ggplot(imput_num_steps, aes(date, steps))
imput_plot <- imput_plot + geom_histogram(stat = "identity", binwidth = bin, colour = "white")
imput_plot <- imput_plot + scale_x_date(breaks = seq(min(imput_num_steps$date)-10,
                                   max(imput_num_steps$date),
                                   bin),
                      labels = date_format("%m/%d/%Y"),
                      limits = c(as.Date("2012-10-01"),
                                 as.Date("2012-11-30")))
imput_plot <- imput_plot + theme_bw() + xlab(NULL) + theme(axis.text.x  = element_text(angle=45,
                                                                     hjust = 1,
                                                                     vjust = 1))
imput_plot <- imput_plot + labs(title = "Total Number of Steps Taken per Day Using Imputed Data")
imput_plot

Calculate and report the mean and median of the total number of steps taken each day.

cat("The mean of the total number of steps is ", mean(imput_num_steps$steps, na.rm = TRUE))
## The mean of the total number of steps is  10766.19

cat("The median of the total number of steps is ", median(imput_num_steps$steps, na.rm = TRUE))
## The median of the total number of steps is  10766.19

Compare values between the raw data set and the imputed data set.

cat("The mean of the total number of steps in the raw data set is ", mean(num_steps$steps, na.rm = TRUE))
## The mean of the total number of steps in the raw data set is  9354.23

cat("The mean of the total number of steps in the imputed data set is ", mean(imput_num_steps$steps, na.rm = TRUE))
## The mean of the total number of steps in the imputed data set is  10766.19

cat("The median of the total number of steps in the raw data set is ", median(num_steps$steps, na.rm = TRUE))
## The median of the total number of steps in the raw data set is  10395

cat("The median of the total number of steps in the imputed data set is ", median(imput_num_steps$steps, na.rm = TRUE))
## The median of the total number of steps in the imputed data set is  10766.19

Determine if there are differences in activity patterns between weekdays and weekends.

imput_table$date <- as.Date(imput_table$date)
imput_table$day <- weekdays(imput_table$date)
imput_table$day <- ifelse(imput_table$day %in% c("Saturday", "Sunday"), "Weekend", "Weekday")
imput_table <- transform(imput_table, day = factor(day))
xyplot(steps ~ interval | day, data = imput_table, type = "l", layout = c(1,2))