PA1_template.md

project1

Aiqun Huang

Loading and processing data

unzip("activity.zip")
data <- read.csv("activity.csv", header=TRUE)
data$date <- as.Date(data$date)
data_n <- na.omit(data)

What is mean total number of steps taken per day?

Sum the steps taken per day and plot the histgram:

daily <- tapply(data_n$steps, data_n$date, sum, simplify=T)
hist(x=daily, breaks=20, col="red", xlab="total steps per day")

The mean and median are :

mean(daily)

## [1] 10766.19

median(daily)

## [1] 10765

What is the average daily activity patter

avgSteps <- aggregate(data_n$steps, list(interval = as.numeric(as.character(data_n$interval))), FUN = "mean")
names(avgSteps)[2] <- "steps"
with(avgSteps,
     plot(interval,
          steps,
          type="l",
          xlab="interval index",
          ylab="average steps in the interval"))

Which 5-minute interval, on average across all the days in the dataset, contains the maximum number of steps?

avgSteps[avgSteps$steps == max(avgSteps$steps), ]

##     interval    steps
## 104      835 206.1698

Imputing missing values

the total number of missing values in the dataset

sum(is.na(data))

## [1] 2304

Replace the missing value by the mean for that interval

new_data <- data
nas <- is.na(new_data$steps)
avg <- tapply(new_data$steps, new_data$interval, mean, na.rm=TRUE, simplify=T)
new_data$steps[nas] <- avg[as.character(new_data$interval[nas])]

Making a histogram

new_daily <- tapply(new_data$steps, new_data$date, sum, simplify=T)
hist(x=new_daily, breaks=20, col="red", xlab="total steps per day", main="Missing values replaced by mean")

The new mean and median are

mean(new_daily)

## [1] 10766.19

median(new_daily)

## [1] 10766.19

After inserting the missing values by the mean, the new mean does't change, the new median is 10766.19 compared to the original 10765.

Are there differences in activity patterns between weekdays and weekends?

Create a factor varible wk with two levels weekday and weekend, the commented code also works.

is_weekday <- function(d) {
    wd <- weekdays(d)
    ifelse (wd == "Saturday" | wd == "Sunday", "weekend", "weekday")
}
#new_data$weekdays <- factor(format(new_data$date, "%A"))
#levels(new_data$weekdays)
#levels(new_data$weekdays) <- list(weekday = c("Monday", "Tuesday",
#                                             "Wednesday", 
#                                             "Thursday", "Friday"),
#                                 weekend = c("Saturday", "Sunday"))
wx <- sapply(new_data$date, is_weekday)
new_data$wk <- as.factor(wx)

Average the steps for all intervals in weekday and weekend, and make the plot.

wk_df <- aggregate(new_data$steps, 
                      list(interval = as.numeric(as.character(new_data$interval)), 
                           wk = new_data$wk),
                      FUN = "mean")
#wk_df <- aggregate(steps ~ wk+interval, data=new_data, FUN=mean)
names(wk_df)[3] <- "steps"
library(lattice)
xyplot(steps ~ interval | wk,
       layout = c(1, 2),
       xlab="Interval",
       ylab="Number of steps",
       type="l",
       lty=1,
       data=wk_df)