Fitbit activity data explorer

Andrew Weston
May 22, 2016

Description

The Fitbit activity data explorer is a web application built using R and the Shiny package. Its purpose is to display activity and movement data collected by a Fitbit personal fitness tracker in an interactive data frame.

This is part of the Coursera Data Science Developing Data Products course.

Data source

For demonstration purposes, my own personal Fitbit data was used. This was obtained from the fitbit website at http://www.fitbit.com/ via the “export data” function.

In the future, the ability to upload other fitbit data (e.g. your own personal data) will be added.

source("fitbitdata.R")
str(fb)
'data.frame':   93 obs. of  10 variables:
 $ Date                  : chr  "2/20/2016" "2/21/2016" "2/22/2016" "2/23/2016" ...
 $ Calories.Burned       : num  1940 2025 2219 2371 2079 ...
 $ Steps                 : num  1923 2650 5517 7391 3859 ...
 $ Distance              : num  0.86 1.18 2.46 3.29 1.72 3.05 4.02 1.52 0.29 3.47 ...
 $ Floors                : int  15 10 13 22 14 22 34 12 1 23 ...
 $ Minutes.Sedentary     : num  867 930 677 655 786 ...
 $ Minutes.Lightly.Active: num  84 117 158 172 113 154 267 120 27 215 ...
 $ Minutes.Fairly.Active : int  0 0 6 6 0 14 2 0 0 5 ...
 $ Minutes.Very.Active   : num  0 0 3 11 0 9 9 0 0 4 ...
 $ Activity.Calories     : num  303 436 693 851 492 ...

Demonstration

The app itself can be accessed at https://saladami.shinyapps.io/fitbit/

The code can be accessed on github at https://github.com/saladami/developing-data-products

shinyServer(function(input, output) 
{
    output$contents <- 
        renderTable({fb[input$Date.range[1]:input$Date.range[2],input$variable]})

})

Improvements

Ideas for improvement (i.e. what I should have done given more time):

  • support for sorting by column
  • charts / fancy plots / trends
  • heart rate / weight / body fat data