#100daysofqs: Daily Art From Data
After 10 years of collecting data on herself, data visualizer Lillian Karabaic embarked on a project to make a daily art piece from her data for 100 consecutive days, with pieces ranging from "Mildly Scary Things I Have Done" to "Burritos Per Year."
Google sheet | MyFitnessPal | SleepCycle | UP Coffee
100 Days of QS Daily Art from DataHi, I’m Lillian. I like regressions and long days to walk on the beach. When I was 20 years old in 2017, I began tracking statistics about myself in a giant Excel spreadsheet. I kind of didn’t have a plan. I just started doing this, and I have really no idea why I started doing it. I just started doing it one day, and once I started it was really hard to stop.
So, I did a talk last year at QS17 about all the tools I use, so I won’t get too deep into that because that is a whole talk by itself. The iPhone kind of gave my neurotic data driven self a lot of tools to collect data and I use those tools to make a bunch of art. So, if you want to find out the actual tools I use, you can go that, my life stack on my website and I update that pretty regularly.
So, each year I take all the data I collect, and I turn it into an annual report. And I do this like in a marathon session usually of like 48 hours, I lock myself in my room and then I try to find interesting things to say about the data.
I’m an economist so we can turn anything into a story about data because we’re trained in lying with data.
So these are a collection of some of the zines that I made, and I send these all over the world, so quite a few people in this room are probably on the list of people that get it. But every year I send the zine out to the world. But last year, I was kind of bad. I made a book instead and I didn’t write an annual report.
And I’m a mid-westerner so I’m driven by guilt and I felt really bad about not making an annual report last year. Even although you know, making a 260-page book and laying it out and printing it and shipping it all over the world kind of takes a lot of work. Fun fact.
And so, the book went to the printer in late March, and I heard about this thing called The 100 Day Project, which is a thing on Instagram where people try to do something creative everyday and post about it. And lots of incredibly talented visual artists do it and just really accomplish really artistic people, none of which I am. So, but I like to pretend I am, so I decided to imitate them and try it.
I also liked it because I’m not really one to make life easy for myself. Pretty much I will always turn something into a project if possible, in order to punish myself in the most efficient way possible. So, the whole project ran from April 4 to July 11, which was kind of nice because it led pretty close into my birthday.
So, let’s just dive into the art because a hundred pieces of art is way too many to show you in seven and a half minutes. But I will show you some of the highlights and lowlights from what I learned about this.
So, I made my first mistake the very first day. I was used to making these huge annual reports and I usually will try to find interesting things to do and then I will work really hard on them. I didn’t work with one of my existing datasets on the very first day. This is all of the states I went to by year, every single year of the past decade and I pulled that data from my email history. Three different forms of location data. Tour posters from bands I toured with, an archive of plane and train tickets, and used it to create a dataset and had to lay it out in Illustrator.
And I used to travel by trains exclusively, so there were a whole bunch of states. In the end I got to this piece which is cool but it’s not even that cool of a color scale because I thought, maybe I want to do greyscale and I gave up after the first day. And fun fact. This took three hours and 22 minutes to put together. I work two jobs. I run a business. I’ve a radio show. I can’t spend 21 hours a week making data for this entire project. So, I clearly had to figure out a better way to do this.
So, first lesson, last big project.
One thing I learned is that the type of data and what it looks like, how much data is in an image has literally nothing to do with how long it took to create. So, all the images up there will show you how long they took to create.
The one on this side, the coffee image, that came from a CSV dataset. All I had to do was export it. It took almost no time to make. I used a template that I had like create.
On this side, this is my Flicker dataset. Flicker has terrible search algorithms. Fun fact. So, I had to dig through 10,000 different photos over 01 years in order to pull this, and my email receipts to figure out when my cameras were bought.
In the end I’m not even happy with that image, but it took so much longer to create. CSVs are awesome.
So in the beginning I tried to cram a ton of information, and into one photo, you have to get in into the Instagram’s square image. So, as you can see, I tried to cram like all of this information into one photo.
I have 10 years of data on my booze, caffeine, alcohol, sleep. I started adding data and I never stopped. So, I have really large datasets. I discovered I could fit 10 years into one slide. It just wasn’t possible.
So, I started to get a little more ridiculous with what I did. I found that these actually the simplest post ended up performing best. People were most interested like in these weird random statistical correlations that I found in my data.
I also cheated a few times, and repurposed existing slides that I had. So, the slide on the left is from my anomaly dot world flight, which was a trip that I took a year ago almost today that had a live satellite feed of where I was at. So, all I did was mash that up with my pages of books that I have read.
And then the project on the other side was actually from a freelance piece that I had that was published during the project, and that was one of the ones that they didn’t have the rights to because they didn’t end up using it. So, I was like, this is an easy way to do promo. So, the reason it says it takes 10 hours to put together on the top is because it was from a huge freelance piece that I did on my entire wardrobe. But in the end, that was the only one I published so that actually only took 10 minutes on the day of.
Things I learned, 100 days is a long time. People don’t pay that much attention to Instagram, so I could use the same data sources over and over again.
My coffee was a CSV, so I just kept repurposing the same data sources. This is the exact same data source. On one side, I tried an Instagram style you know, click-baity, fancy flat-laid post, and then on this side I kind of based on Nicholas Felton’s annual reports, which if you like any of this you’ll like his stuff better because he’s a real designer.
I also found that this kind of style where I use a photo and then just put a small amount of data on it did really well. They were kind of slices of life with small amounts of data. But I also learned that 100 days is a long time to not take a day off. So, I had a couple of days where I really stretched the idea of what a graph was.
On one side, that was my one day that I actually didn’t work at all and I went to the coast and I made a graph of the days that I had gone to the coast in the past couple of years in the sand. And then on the other side I had spent 14 hours shipping books every day for three days straight, and I was like, what do I have available to make a graph.
I also just kind of lost motivation at some point. This was on day 28 when I realized I had still 72 days to go. And then I would start to get a little surreal. I can’t explain either of these. I kind of just decided what was a graph, I don’t know.
And in fact, as the project went on, I got even a little lazier, which was my real saving grace was Exist.IO, which I don’t think they’re here because they’re Australians, but they are a data aggregator that will pull all your data sources. And they do a really excellent job of design, and so a few days or I did was take screenshots and then put them in my fonts.
I also found that food -related posts did really well. Like neither of these are particularly interesting datasets, but if you put pictures of food on Instagram people love it.
You can also see my life events over the hundred days reflected in this. So, one of these was the day that I published the book, and the other was the day that I got my bike worked on for the first time in three years. Don’t do that.
I also updated some kind of existing projects. As the project was getting close to the end I kind of remembered that the whole point of this was to create a zine, so I updated some old slides from my previous zine.
And I got really stressed out towards the last couple of days. This was day 95 when I was like, ‘Oh I have all this data. I should do something with it’. Look how long that took. Also, that’s only one slide. I actually made a lot more graphs on that day and I put them on the slide through on Instagram.
In the end, I got to my birthday, and it was actually a kind of a letdown because my birthday was the seventh and the last day of the project was the eleventh. So then I just had to just figure out what to do for the last couple of days. But this was the kind of final slide, which of course, I want to spend my birthday dressed in this silly costume, compiling data all day.
Because this is Quantified Self, I figured you wanted the aggregate data. This is my hours per week. This really reflects a kind of an inverse of working on my other jobs. It’s almost, I wish I could have mapped it more easily for you. But pretty much when I didn’t have a lot of work to do, I would spend a lot of time on this project.
So, what did I learn?
A hundred days is a long time! I’ve did a lot of 365 projects, where you take a photo every day but it turns out doing a bunch of design work every day is a lot more work. If you work at anything consistently you will just get lazier, because you will get faster and you will know how to take shortcuts. Consistent data formats are amazing. Pretty much I had to work with a CSV, I was so happy. And every time I had to like take a GPS and then overlay a CSV onto it, and then put it into Illustrator, my whole time went a lot longer.
I actually spent 5% of the total time on this project, just dealing with uploading to Instagram, which you can only do from mobile, which involved me sending a lot of iPhone message pictures to myself. And, quite often the data did not serve the square format, and that was a really frustrating part of Instagram which was trying to work within that constraint.
So, as I mentioned a few hundred times, I had a book come out this year. If you want to get the book its (Cat’s Plane? 10:55) Personal Finance, come and find me and as I mentioned I will turn this into a zine. And I send the zine’s all over the world, so if you just go to that kind of simple website, you put in your email I’ll alert you when the zine comes out. Or come find me and I will sell them to you.