Personal Movement Monitoring Devices

Introduction

It is now possible to collect a large amount of data about personal movement using activity monitoring devices such as a Fitbit, Nike Fuelband, or Jawbone Up. These type of devices are part of the โ€œquantified selfโ€ movement โ€“ a group of enthusiasts who take measurements about themselves regularly to improve their health, to find patterns in their behavior, or because they are tech geeks. But these data remain under-utilized both because the raw data are hard to obtain and there is a lack of statistical methods and software for processing and interpreting the data.

This analysis makes use of data from a personal activity monitoring device. This device collects data at 5 minute intervals through-out the day. The data consists of two months of data from an anonymous individual collected during the months of October and November, 2012 and include the number of steps taken in 5 minute intervals each day.

The data for this analysis can be downloaded from the course web site:

The variables included in this dataset are:

  • steps: Number of steps taking in a 5-minute interval (missing values are coded as ๐™ฝ๐™ฐ)
  • date: The date on which the measurement was taken in YYYY-MM-DD format
  • interval: Identifier for the 5-minute interval in which measurement was taken

The dataset is stored in a comma-separated-value (CSV) file and there are a total of 17,568 observations in this dataset.

Loading and preprocessing the data

Unzip data to obtain a csv file.

fileUrl <- "https://d396qusza40orc.cloudfront.net/repdata%2Fdata%2Factivity.zip"
download.file(fileUrl, destfile = paste0(getwd(), '/repdata%2Fdata%2Factivity.zip'), method = "curl")
unzip("repdata%2Fdata%2Factivity.zip",exdir = "data")

Reading csv Data into Data.Table.

activityDT <- data.table::fread(input = "data/activity.csv")

What is mean total number of steps taken per day?

  1. Calculate the total number of steps taken per day.
Total_Steps <- activityDT[, c(lapply(.SD, sum, na.rm = FALSE)), .SDcols = c("steps"), by = .(date)] 

head(Total_Steps, 10)
          date steps
 1: 2012-10-01    NA
 2: 2012-10-02   126
 3: 2012-10-03 11352
 4: 2012-10-04 12116
 5: 2012-10-05 13294
 6: 2012-10-06 15420
 7: 2012-10-07 11015
 8: 2012-10-08    NA
 9: 2012-10-09 12811
10: 2012-10-10  9900
  1. Make a histogram of the total number of steps taken each day.
ggplot(Total_Steps, aes(x = steps)) +
    geom_histogram(fill = "blue", binwidth = 1000) +
    labs(title = "Daily Steps", x = "Steps", y = "Frequency") +
  theme_classic()
Warning: Removed 8 rows containing non-finite values (stat_bin).

  1. Calculate and report the mean and median of the total number of steps taken per day.
Total_Steps[, .(Mean_Steps = mean(steps, na.rm = TRUE), Median_Steps = median(steps, na.rm = TRUE))]
   Mean_Steps Median_Steps
1:   10766.19        10765

What is the average daily activity pattern?

  1. Make a time series plot (i.e.ย ๐š๐šข๐š™๐šŽ = โ€œ๐š•โ€) of the 5-minute interval (x-axis) and the average number of steps taken, averaged across all days (y-axis).
IntervalDT <- activityDT[, c(lapply(.SD, mean, na.rm = TRUE)), .SDcols = c("steps"), by = .(interval)] 

ggplot(IntervalDT, aes(x = interval , y = steps)) + geom_line(color="blue", size=1) + labs(title = "Avg. Daily Steps", x = "Interval", y = "Avg. Steps per day") +
  theme_classic()

  1. Which 5-minute interval, on average across all the days in the dataset, contains the maximum number of steps?
IntervalDT[steps == max(steps), .(max_interval = interval)]
   max_interval
1:          835

Imputing missing values

  1. Calculate and report the total number of missing values in the dataset (i.e.ย the total number of rows with ๐™ฝ๐™ฐs)
activityDT[is.na(steps), .N ]
[1] 2304
# alternative solution
nrow(activityDT[is.na(steps),])
[1] 2304
  1. Devise a strategy for filling in all of the missing values in the dataset. The strategy does not need to be sophisticated. For example, you could use the mean/median for that day, or the mean for that 5-minute interval, etc.

Filling in missing values with median of dataset.

activityDT[is.na(steps), "steps"] <- activityDT[, c(lapply(.SD, median, na.rm = TRUE)), .SDcols = c("steps")]
  1. Create a new dataset that is equal to the original dataset, but with the missing data filled in.
data.table::fwrite(x = activityDT, file = "data/tidyData.csv", quote = FALSE)
  1. Make a histogram of the total number of steps taken each day and calculate and report the mean and median total number of steps taken per day. Do these values differ from the estimates from the first part of the assignment? What is the impact of imputing missing data on the estimates of the total daily number of steps?
# total number of steps taken per day
Total_Steps <- activityDT[, c(lapply(.SD, sum)), .SDcols = c("steps"), by = .(date)] 

# mean and median total number of steps taken per day
Total_Steps[, .(Mean_Steps = mean(steps), Median_Steps = median(steps))]
   Mean_Steps Median_Steps
1:    9354.23        10395
ggplot(Total_Steps, aes(x = steps)) + geom_histogram(fill = "blue", binwidth = 1000) + labs(title = "Daily Steps", x = "Steps", y = "Frequency") +
  theme_classic()

Type of Estimate | Mean_Steps | Median_Steps

First Part (with na) | 10765 | 10765 Second Part (fillin in na with median) | 9354.23 | 10395

Are there differences in activity patterns between weekdays and weekends?

  1. Create a new factor variable in the dataset with two levels โ€“ โ€œweekdayโ€ and โ€œweekendโ€ indicating whether a given date is a weekday or weekend day.

Just recreating activityDT from scratch then making the new factor variable. (No need to, just want to be clear on what the entire process is.)

activityDT <- data.table::fread(input = "data/activity.csv")
activityDT[, date := as.POSIXct(date, format = "%Y-%m-%d")]
activityDT[, `Day of Week`:= weekdays(x = date)]
activityDT[grepl(pattern = "Monday|Tuesday|Wednesday|Thursday|Friday", x = `Day of Week`), "weekday or weekend"] <- "weekday"
activityDT[grepl(pattern = "Saturday|Sunday", x = `Day of Week`), "weekday or weekend"] <- "weekend"
activityDT[, `weekday or weekend` := as.factor(`weekday or weekend`)]
head(activityDT, 10)
    steps       date interval Day of Week weekday or weekend
 1:    NA 2012-10-01        0      Monday            weekday
 2:    NA 2012-10-01        5      Monday            weekday
 3:    NA 2012-10-01       10      Monday            weekday
 4:    NA 2012-10-01       15      Monday            weekday
 5:    NA 2012-10-01       20      Monday            weekday
 6:    NA 2012-10-01       25      Monday            weekday
 7:    NA 2012-10-01       30      Monday            weekday
 8:    NA 2012-10-01       35      Monday            weekday
 9:    NA 2012-10-01       40      Monday            weekday
10:    NA 2012-10-01       45      Monday            weekday
  1. Make a panel plot containing a time series plot (i.e.ย ๐š๐šข๐š™๐šŽ = โ€œ๐š•โ€) of the 5-minute interval (x-axis) and the average number of steps taken, averaged across all weekday, days or weekend days (y-axis).
activityDT[is.na(steps), "steps"] <- activityDT[, c(lapply(.SD, median, na.rm = TRUE)), .SDcols = c("steps")]
IntervalDT <- activityDT[, c(lapply(.SD, mean, na.rm = TRUE)), .SDcols = c("steps"), by = .(interval, `weekday or weekend`)] 

ggplot(IntervalDT , aes(x = interval , y = steps, color=`weekday or weekend`)) + geom_line() + labs(title = "Avg. Daily Steps by Weektype", x = "Interval", y = "No. of Steps") + facet_wrap(~`weekday or weekend` , ncol = 1, nrow=2) +
  theme_classic()

Avatar
Rihad Variawa
Data Scientist

I am the Sr. Data Scientist at Malastare AI and head of global Fintech Research, responsible for overall vision and strategy, investment priorities and offering development. Working in the financial services industry, helping clients adopt new technologies that can transform the way they transact and engage with their customers. I am passionate about data science, super inquisitive and challenge seeker; looking at everything through a lens of numbers and problem-solver at the core. From understanding a business problem to collecting and visualizing data, until the stage of prototyping, fine-tuning and deploying models to real-world applications, I find the fulfillment of tackling challenges to solve complex problems using data.

Next
Previous