PA1_template.Rmd

#Call the libraries

library(tidyverse)
library(mice)

#Read the downloaded data from the desktop

activity <- read.csv("C:/Users/zterjek/Desktop/activity.csv")

#Calculate the total steps taken per day, visualize it with a histogram

perdaysum <- aggregate(steps ~ date, activity, sum)

hist(perdaysum$steps, breaks = 20, col = "steelblue")

#Calculate and report the average and median of total taken steps

perdaymean <- mean(perdaysum$steps)

perdaymed <- median(perdaysum$steps)

perdaymean
perdaymed

#Calculate the average steps taken by intervals, visualize it with a line chart

perintervalmean <- aggregate(steps ~ interval, activity, mean)

plot(perintervalmean$interval, perintervalmean$steps, type = "l", col = "darkred",
     lwd = 1.5)

#Calculate and report the interval when max steps were taken

maxstepsperinterval <- max(perintervalmean$steps)

maxstepsinterval <- filter(perintervalmean, steps == maxstepsperinterval)

maxstepsinterval$interval

#Calculate the missing values

md.pattern(activity)

#Impute the missing values with predictive mean matching

activity_imputed <- data.frame(original = activity$steps,
    imputed_steps = complete(mice(activity, method = "pmm"))$steps)

activity_imputed$num <- 1:17568
activity$num <- 1:17568

merged_activity <- merge(activity, activity_imputed, by = "num")
merged_activity <- merged_activity[,-c(1, 2)]
merged_activity <- rename(merged_activity, original_steps = original)
md.pattern(merged_activity)

#Calculate the total steps taken per day of the imputed dataset, 
#visualize it with a histogram

perdaysum_merged <- aggregate(imputed_steps ~ date, merged_activity, sum)

hist(perdaysum_merged$imputed_steps, breaks = 20, col = "darkgreen")

#Calculate and report the average and median of total taken steps in the 
#imputed dataset

perdaymean_merged <- mean(perdaysum_merged$imputed_steps)

perdaymed_merged <- median(perdaysum_merged$imputed_steps)

perdaymean_merged
perdaymed_merged

#Checking the weekdays

merged_activity$date <- as.Date(merged_activity$date)

merged_activity$daytype <- as.factor(ifelse(weekdays(merged_activity$date) %in% 
                        c("szombat", "vasárnap"), "Weekend", "Weekday"))

#Calculate the average steps taken by intervals, visualize it with a line chart 
#with weekdays and weekends separated

grouped <- merged_activity %>% group_by(daytype, interval) %>%
    summarize_at("imputed_steps", mean)

ggplot(grouped, aes(interval, imputed_steps))+
    geom_line(aes(col = daytype))+
    facet_wrap(.~ grouped$daytype, nrow = 2, ncol = 1)