As I was finishing my Synthesis 2 assignment, the biggest regret was that I ended up visualizing my data in a much more conventional manner and didn’t end up learning anything new. There was just such a large gap in my conceived product and the actual product that I know could only be made up by a year or two of practice and experience… which I didn’t have. Hence, the group project appealed to me because I knew everyone in my group had more experience than me and I wanted to take the chance to see the strategies they use to problem solve. This includes anything as broad as Pandas to small details like how often do they refresh their html page to check if their changes were right?
It seems like everyone else had the same idea. We began the brainstorming process by ruling out the project ideas that would utilize things we have already done. All of us, except for Jahari, have not yet used maps for our synthesis assignments because it was so daunting to do alone, but want to pursue now that it’s a group project. Thus, Jahari unwittingly became the anchor of this project.
Because he has had experience, we wanted to use his code from his earlier project and simply input another dataset and make the visualization more interactive than simply visualization. Because Maya used a slider to compare her data in chronological order, we chose that element to be our interactivity. When none of the datasets online really resonated with us, I suggested the only think I know how to do, which is generating/modifying our own data sets.
I thought it would be cool to have a data-driven diary visualized, and compare our lives to each other as Pitt students but with different majors. The most effortless method to track an event would be a GPS app that tracks where we frequent. Hence, we were asking the question, “Where do Pitt students go, and how do those places differ due to our studies?” My (only) contribution was finding apps for both iPhones and Android that generated the data we want. Apps that had no method of exporting the data were ruled out, as well as apps that delete location history everyday.
Our data spans a week so both our weekday and weekend lives are incorporated. It was plotted (mostly by Jahari) on a Bokeh map that has Google Maps overlayed so we can recognize landmarks. Connor created a website template and miraculously embedded the map onto our homepage. Maya refined that webpage so everything was aesthetically pleasing. Kat organized the data into a csv file and plotted it on My Maps to get alternative visualization of not our locations but our routes, something the bokeh map cannot express with simply dots. It can be accessed here in each of the headers under our individual page.
Some problems we encountered were how the android apps had a higher sampling rate than the iPhone apps. However, that didn’t need to be solved. The google map could not shaved to fit the bokeh frame. Googling did not help this solution. There was also the problem of embedding the bokeh map into html, but I think Javascript solved that problem. Finally, Kat and I wanted to visualize our analysis of our routes by plotting how frequent we visit Cathy, WPU, or Market on piecharts or bar graphs but we realized there was no automated way of finding those coordinates, so that idea was dropped.
While I was not able to contribute too much to the coding aspect of this project, I learned a lot just by watching how everyone worked through problems, researched, and decide whether to continue or drop an idea. Because for this project I did not have enough skills to offer, I’m thinking I can apply the skills I learned in the final project instead.
Jahari’s routes (in purple) appeared oddly spiky, and upon taking a closer look at the time stamps and location coordinates, it is revealed that he has been teleporting to Schenley Park at night. Similar incidents also happened to the rest of us, like Maya swimming in the Allegheny River, Connor zipping in a perfect line to New York, or a mysterious visit to the airport. The GPS is probably pinging inconsistently to different cell towers, resulting in these GPS points marking locations we were not at. We originally wanted to clean our data sets but decided to leave it all in to show that flaw.
The feeling of surveillance was also felt when we looked at the final visualization. We were only aiming to analyze where we attend class, and obtain a lighthearted look at our weekly lives. However, the amount of detail the GPS gave us revealed more intimate details, such as our exact apartment, the classroom we sit in, where we work, where we eat. We could deduce Connor traveled to New York via private shuttle because he has not visited the airport nor the Megabus and Greyhound bus stops. People asked me what I was doing on Wylie Ave at midnight (work). In addition, because of Google Maps, we were able to recognize landmarks, such as all the Asian restaurants Connor frequents to eat, Kat practically living in the Cathedral, and Jahari’s study spot at Benedum. Bus routes were also visualized. Connor and Jahari live on upper campus, so the 10A route is roughly highlighted. The 71 and 61 bus routes from campus to downtown also shows up through my frequent travels to work. Overall, it looks like us students don’t leave campus much, and if we do, we can only travel to the places where there are public transportation.
I also want to stress that this is not an accurate picture either. Some things we neglected to consider were all the times our phone went dead, on airplane mode, or underground. For example, when I take the T to work, my phone will not be tracked because I am underground the whole time. But despite these details, I feel very proud that our project was created completely from scratch, and how the resulting product gives a personal look to our daily lives.