PA1_template.Rmd

# Reproducible Research: Peer Assessment 1

```{r,echo=TRUE}
require(ggplot2)
require(timeDate)
```

## Loading and preprocessing the data
```{r}
data<-read.csv('activity.csv')
data$intfact <- as.factor(data$interval)
data3<-data
data3$date<-as.Date(strptime(data3$date,"%Y-%m-%d"))
data3$weekdays<-isWeekday(data3$date,wday=1:5)
```


## What is mean total number of steps taken per day?

### mean:

```{r,echo=TRUE}
y<-aggregate(data$steps,by=list(data$date),FUN=mean)
y$x
qplot(Group.1,data=y,weight=x,geom="histogram",ylab="mean",xlab="date")
```

### median:

```{r,echo=TRUE}
y<-aggregate(data$steps,by=list(data$date),FUN=median,na.action=na.omit)
y$x
qplot(Group.1,data=y,weight=x,geom="histogram",ylab="median",xlab="date")
```

## What is the average daily activity pattern?

### daily activity:

```{r,echo=TRUE}
data2<-data[!is.na(data$steps),]
y<-aggregate(data2$steps,by=list(data2$intfact),FUN=mean)
y$x
plot(y$x~y$Group.1,type="l",ylab="mean",xlab="intervals")
```

### interval with maximum number of steps:

```{r,echo=TRUE}
y$Group.1[which(y$x == max(y$x))]
```

## Imputing missing values

### number of missing rows: 

```{r,echo=TRUE}
length(data$steps[is.na(data$steps)])
```

### fix data:

remove all NAs and put mean number of steps instead

```{r,echo=TRUE}
data$steps[is.na(data$steps)]<- mean(data$steps,na.rm=TRUE)
```

### calculate mean and median again:

### mean:

```{r,echo=TRUE}
y<-aggregate(data$steps,by=list(data$date),FUN=mean)
y$x
qplot(Group.1,data=y,weight=x,geom="histogram",ylab="mean",xlab="date")
```

### median:

```{r,echo=TRUE}
y<-aggregate(data$steps,by=list(data$date),FUN=median,na.action=na.omit)
y$x
qplot(Group.1,data=y,weight=x,geom="histogram",ylab="median",xlab="date")
```

Mean and median values slightly differ from original data set


## Are there differences in activity patterns between weekdays and weekends?

TRUE = weekday, FALSE = weekend

```{r,echo=TRUE}
data3<-data3[!is.na(data3$steps),]
y<-aggregate(data3$steps[data3$weekdays == TRUE],by=list(data3$intfact[data3$weekdays == TRUE]),FUN=mean)
names(y)<-c("interval","steps")
z<-aggregate(data3$steps[data3$weekdays == FALSE],by=list(data3$intfact[data3$weekdays == FALSE]),FUN=mean)
names(z)<-c("interval","steps")
y$weekdays<-TRUE
z$weekdays<-FALSE
w<-merge(y,z,all=TRUE)
qplot(interval,steps,data=w,facets=weekdays~.)
```