PA1_template.Rmd

---
title: "Reproducible Research: Peer Assessment 1"
output: 
  html_document:
    keep_md: true
---


Dplyr is used during the loading and preprocessing the data


```{r, echo=TRUE}
setwd("C:/Users/Simon/Google Drive/R/coursera/reproducible research/RepData_PeerAssessment1")
unzip("activity.zip")
activity <- read.csv("activity.csv")
suppressPackageStartupMessages(library(dplyr))

```


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

Total steps per day
```{r, echo=TRUE}
activity %>% 
        group_by(date) %>% 
        summarize(total=sum(steps)) %>% 
        print
```


```{r, echo=TRUE}
hist(activity$steps, xlab="Steps", main="Histogram of overall steps")
```

Mean and median number of steps per day
```{r, echo=TRUE}
activity %>% 
        group_by(date) %>% 
        summarize(mean=mean(steps, na.rm=TRUE), median=median(steps, na.rm=TRUE)) %>% 
        print
```


## What is the average daily activity pattern?
```{R echo=TRUE}
interval_orig <-activity %>%
        group_by(interval) %>%
        summarize(meanint=mean(steps, na.rm=T))
order_int <- interval_orig %>%
        arrange(desc(meanint))
        
max_steps <- order_int[[1]][[1]]
        
````
The 5 minute interval with the highest number of steps is `r max_steps`  as seen in the time series plot below.
```{R echo=TRUE}


plot(interval_orig$interval,interval_orig$meanint, type="l", xlab="interval", ylab="Mean steps")
```




## Imputing missing values
```{R, echo=TRUE}
missnum <- activity %>%
        is.na %>%
        sum
````
Nr of missing values are `r missnum`


```{R, echo=TRUE}
missingsteps <- is.na(activity$steps)
activity$steps[missingsteps] <- mean(activity$steps,na.rm=T)

impute_activity <- data.frame(activity)
activity <- read.csv("activity.csv")
````

```{R, echo=TRUE}
total_impute <- impute_activity %>% 
        group_by(date) %>% 
        summarize(total=sum(steps))
hist(total_impute$total,xlab="Total steps", main="Histogram of total steps")
````

Mean & median steps per day
```{R, echo=TRUE}
impute_activity %>% 
        group_by(date) %>% 
        summarize(mean=mean(steps, na.rm=TRUE), median=median(steps, na.rm=TRUE)) %>% 
        print
````

Does these numbers differ from the first part of the assignment? Yes.The imputation increased the proportion of low activity compared to the non-imputed values seen below.
```{R, echo=TRUE}
activity %>% 
        group_by(date) %>% 
        summarize(mean=mean(steps, na.rm=TRUE), median=median(steps, na.rm=TRUE)) %>% 
        print
````

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

```{R, echo=TRUE}
stripped_date <- strptime(impute_activity$date,format="%Y-%m-%d")
daysofweek <- weekdays(stripped_date)
weekends <- daysofweek=="Saturday"|daysofweek=="Sunday"
workdays <- daysofweek=="Monday"|daysofweek=="Tuesday"|daysofweek=="Wednesday"|daysofweek=="Thursday"|daysofweek=="Friday"
weekly_activity <- data.frame(impute_activity, workdays, weekends)
````

```{R, echo=TRUE}
weekend_steps <- weekly_activity[weekends==TRUE,] %>% group_by(interval) %>% summarize(mean=mean(steps,na.rm=T))
workday_steps <- weekly_activity[weekends==FALSE,] %>% group_by(interval) %>% summarize(mean=mean(steps,na.rm=T))
par(mfcol=c(2,1))
plot(workday_steps$interval,workday_steps$mean, type="l", main="Workdays", xlab="Interval", ylab="Number of steps")
plot(weekend_steps$interval,weekend_steps$mean, type="l", main="Weekends", xlab="Interval", ylab="Number of steps")
````