Quantified Self – Analysis of my Time

Almost every day I am looking at data either for my job or personal development. As an analyst, I am always trying to synthesize stories from data to help bring insight to my team. Now I am starting to look at my habits and figure out how I can synthesize stories about myself. Recently I stumbled across the concept of a Quantified self.

Quantifying myself means understanding my behaviors, habits, and unseen quirks through observations made and recorded about those behaviors.

I work pretty far from where I live, so my wife always wants to know when I will be home. Because of that, I built a Tasker app that allowed me to send a geocoded version of my location and an estimated time of how far I was away to her. Google has some great APIs that can help with that. That worked for a while, but because I had an LG Android phone, the OS failed. I had to build my Tasker apps from scratch. Instead of texting my wife my location, I decided to build a small Tasker app that records my location every 15 minutes in a Google Spreadsheet.

She was able to see my current location and how much battery life my phone had left. After that, I got to thinking that if I know my current location, there is a good chance I know what I am doing. For example, if my GPS says I am in Pasadena I am most likely at home.

For Christmas last year, I received a Fitbit from my in-laws. Combining that data and my location data I was recording for my wife; I was able to get a general idea of what I was doing at almost any time of the day.

There are all kinds of analysis I can run on this data, but I started with a basic timeline to know where I am spending my time and what I am doing. Below is the first visual I have created from this data. Over the next couple weeks, I will be writing some tutorials on how I created a location tracker using Google Spreadsheets & Tasker and how I pulled data from FitBit’s API and combined it with location information.

plot of chunk unnamed-chunk-1

I want to develop as many ways possible to collect information passively. Here are a few ideas that I will be looking at in the future.

  1. Transaction records from bank to know when I am eating out.

  2. Using an OBDII Bluetooth Connector to collect information from my car like MPG or average miles per hour.

  3. Using Google Calendar to understand how the work day is spent.


comments powered by Disqus