PA1_template.RMD

#Course project 1

## 1. Let's clean the data set first
```{r}
setwd('D:/heyimeng')
activity <- read.csv('activity.csv', na.strings = '')
activity$date <- as.Date(as.character(activity$date))  ## change the format of date
head(activity, 5)
```

## 2. Total number of steps taken per day

#### 1. Caculate the total number of steps taken per day 
```{r}
library(dplyr)
activity1 <- filter(activity, steps != 'NA')
activity1$steps <- as.numeric(activity1$steps)
activity1$date <- as.factor(activity1$date)
steps_per_day <- tapply(activity1$steps, activity1$date, sum)
print(steps_per_day)
```

#### 2. Make a histgram of total number of steps per day
```{r}
hist(steps_per_day,
     xlab = expression('Steps'), 
     main = 'Histogram of steps per day')
```

#### 3. Caculate the mean and median value of steps per day
```{r}
meanvalue <- mean(steps_per_day)
medianvalue <- median(steps_per_day)
print(meanvalue)
print(medianvalue)
```

## 3. Daily activity pattern

#### 1. Time series plot
```{r}
steps_per_interval <- tapply(activity1$steps, activity1$interval, mean)
plot(unique(activity1$interval), steps_per_interval,
            xlab = 'Interval',
            ylab = 'Steps',
            main = 'Time series plot',
            type = 'l')
```

#### 2. Find the maximum number of steps
```{r}
location <- which(steps_per_interval == max(steps_per_interval))
steps_per_interval[location]
```


## 4. Imputing missing value

#### 1. Calculate and report the total number of missing values in the dataset 
```{r}
NAS <- activity$steps == 'NA'
sum(NAS)
```

#### 2&3. Filling all the missing values(use the mean of 5-minute intervals)
```{r}
actna <- filter(activity, steps == 'NA')
Index<- data.frame()
        for (i in 1:288){
                Index[i,1] <- steps_per_interval[[i]]
        }
cn <- names(steps_per_interval)
meanIndex <- cbind(cn,Index)
names(meanIndex) = c('interval', 'steps')
actfilled <- cbind(meanIndex, actna)
actfilled <- actfilled[, -c(1,3)]
actfilled$interval <- as.integer(actfilled$interval)
activityModeifed <- rbind(actfilled, activity1)
head(activityModeifed)
```

#### 4.Histogram of steps taken per day and mean and median valuue
```{r}
steps_per_day_Modified <- tapply(activityModeifed$steps, activityModeifed$date, sum)
hist(steps_per_day_Modified, main = 'Modified histogram', xlab = 'steps')
mean(steps_per_day_Modified)
median(steps_per_day_Modified)
```

## 5. Differences in activity patterns between weekdays and weekends
#### 1. filter weekdays and weekend
```{r}
Sys.setlocale("LC_TIME", "English") 
DATE <- weekdays(activityModeifed$date)
activityModeifed <- cbind(activityModeifed, DATE)
actweekday <- filter(activityModeifed, DATE != 'Sunday' & DATE != 'Saturday' )
actweekday$DATE <- 'Weekday'
actweekend <- filter(activityModeifed, DATE == 'Sunday' | DATE == 'Saturday')
actweekend$DATE <- 'Weekend'
activityModeifed <- rbind(actweekend, actweekday)
activityModeifed <- arrange(activityModeifed, date)
head(activityModeifed)
```

#### 2. Average steps taken between weekday and weekend
```{r}
steps_per_interval_weekend <- tapply(actweekend$steps, actweekend$interval, mean)
steps_per_interval_weekday <- tapply(actweekday$steps, actweekday$interval, mean)
par(mfcol = c(2,1))
plot(unique(activityModeifed$interval), steps_per_interval_weekend, type = 'l',
     col = 'blue', main = 'weekend', xlab = 'Interval', ylab = 'Steps')
plot(unique(activityModeifed$interval), steps_per_interval_weekday, type = 'l',
     col = 'blue', main = 'weekday', xlab = 'Interval', ylab = 'Steps')