---
    title: "Course5Project1"
    author: "Nevon"
    date: "June 12, 2018"
    output:
     html_document:
            keep_md: yes

    ---

    ```{r setup, include=TRUE}
    knitr::opts_chunk$set(echo = TRUE)
    ```

    1. First we load the walking activity data
    ```{r}
    ##Load Data
    fitness <- read.csv("activity.csv")
            df <- data.frame(fitness)
    ```
    2. Next we will calculate base measure
    ```{r}
    ## Calculations/ Metrics
            stepsbyday <- aggregate(steps ~ date, data = df, sum) ## steps per day
            avgsteps <- mean(stepsbyday$steps)## average steps by day
            median <- median(stepsbyday$steps)## median steps by day
    ```
    3. Following the base measures we plot the average steps by day
    ```{r}
    ##Plot Histogram & Report Figures
    hist(stepsbyday$steps, xlab = "Number of Steps Per Day", main = "Total Steps Per Day", breaks = 4, col =              "royal blue")
            ## Add Metrics
            abline(v = median(stepsbyday$steps), col = "red", lwd = 10)
            abline(v = mean(stepsbyday$steps), col = "yellow", lwd = 2)
            legend(x = "topright", c("Median", "Mean"), col =c("red", "yellow"), lwd = c(2, 2, 2 ))
    ```
    4. Afterwards we will look at the steps per intervals 4a.First removing NAs by creating new data set than             plotting the figures
    ```{r}
    ## Calculate Steps by Interval
            library(ggplot2)
            Intervals <- df[!is.na(df$steps), ] ##remove NAs
            intrv <- aggregate(steps ~ interval, data = Intervals, mean)
                    ## Create Plot
                    g <- ggplot(intrv, aes(x=intrv$interval, y = intrv$steps), xlab = "Intervals", ylab = "Avg            Steps")
     g+geom_line() + xlab("Intervals") + ylab("Avg Steps") + ggtitle("Avg number of Stepgs by Intervals")
                            ##Find Max Step Interval
                            max <- max(intrv)
                            print(max)
    ```
    5. Then we calculate the weight of missing values, i.e(how many missing values are there)
    5a. Also we will replace missing values using the value of the average steps per day we calculated before &                create clean data set
    5b. Furthermore we will create a new data set which merges orignal data set with new clean data
    5c. Lastly we will plot the new set & find new measure

    ```{r}
    ## Calculate Weight of Missing Values
                    ALLNAs <- as.numeric(is.na(df))
                    Missing_Val <- sum(ALLNAs)
                    print(Missing_Val)
                            ##Substitute NAs with average steps per date
                            library(plyr)
                            dfvalues <- Intervals
                            avgsteps_day <- tapply(Intervals$steps, Intervals$interval, mean, na.rm = TRUE,                                       simplify = T)
                            NAdata <- is.na(dfvalues$steps)
                            dfvalues$steps[NAdata] <- avgsteps_day[as.character(dfvalues$interval[NAdata])]

    newstepstotal <- tapply(dfvalues$steps,dfvalues$date, sum, na.rm = TRUE, simplify = T) ## New data Frame
            newstepstotal <- newstepstotal[!is.na(newstepstotal)]
    ##Plot New Hist & Find New Metrics
                            hist(x = newstepstotal,
                                    col = "royal blue",
                                    breaks = 10,
                                    xlab = "Daily Steps")

                            ##Find Metrics of Newsteptotal
                            summary(newstepstotal)
    ```
    6.The last part of our anlysis will be to explore differences between weekend & weekdays
                    6a. First create new variable for weekends/weekdays
                    6b. Next find value of steps per daytype
                    6c. Lastly we'll plot the data
    ```{r}
    ## Segment Data into Weekdays / Weekends
            wd <- !(weekdays(as.Date(df$date)) %in% c("Saturday", "Sunday"))
            wknd <- c("","")
            for (i in 1:length(wd)) {
                    if(wd[i]) {wknd[i] <- "Weekday"} else {wknd[i] <- "Weekend"}
            }
            df[, "dayType"] <- factor(wknd) ##new daytpe variable

            wk_df <- aggregate(steps~dayType+interval, data = df, mean)##average steps per daytype
            library(lattice)
            xyplot(steps ~ interval | factor(dayType),
                   layout = c(1,2),
                   xlab = "Interval",
                   ylab = "Number of Steps",
                   type = "l",
                   lty=1,
                   data = wk_df)
    ```