Pedro M Cruz's master thesis
Getting back to the Lisbon only dataset, I have 1534 taxis to map during October 2009. How will it look like using the same production scheme as the last artifact? Eye candy, eye candy. Not much to extract. Also, I’m noticing that I have to go into more complex data filtering and clustering schemes, as the taxis’ routes seem to climb some walls…
When you start in a field, it’s not a bad idea to start imitating some reference work. Aaron Koblin’s Flight Patterns are a reference in information aesthetics. Is this work going to evolve in something closer to Koblin’s Flight Patterns? I plan to differentiate, but I don’t mind to pass through it as natural work path.
Continuing to explore. The dataset with more than 3 million entries for two months, “only” reports to 496 unique taxis. Each one is represented by a single random color. I see how much the metropolitan areas are colorful comparing to the rest of the territory. I also notice how certain routes are stuck to one color, meaning that taxi drivers operate at a regional level (uuh, what an elaborate conclusion…). Looking further.
Looks like I couldn’t avoid myself in taking a look into traffic data (taxis) for all over Portugal, between November and December 2008.
I started exploring some data relative to taxi traffic in Lisbon. Totaling more than 2 million entries, the data reports to October 2009, and for each information received from a taxi, I draw a point in its location.
There is also some data reporting to November and December 2008, of more than 3 million entries. It seems to result in a more organic look that much pleases me. Nevertheless, the data has 2 years old and is more scattered along time and space, which could be good, but as I’m focusing in Lisbon it traduces in a smaller data density. I’m gonna stick with the first dataset for now. Experimenting. How sweet is that I can identify taxi stands?
The actual theme for my master thesis is coming closer to be defined. For now I can tell that my future work path will be something about visualizing books content — very exciting! Having that, I’ve decided to start experimenting on text analysis. As a starting point, the analysis itself was very raw — I limited myself to analyze word frequency in the text.
The text chosen was a very well know Portuguese epic poem — Os Lusíadas. I chose this poem mainly as a provocative towards what it seems a banalized intellectual status among the inhabitants of the Kingdom of Portugal. The result was a collection of 10 static pieces about each of the 10 most frequent words in the poem. All the work was done in Processing, so it wouldn’t be hard to think in some kind of interaction. Although, the main purpose as I stated was to start analyzing text. I wanted to keep the graphical output as simple and elegant as my knowledge allowed.
Os Lusíadas is a Portuguese epic poem by Luís Vaz de Camões first printed in 1572.
The poem consists of ten cantos and 1102 stanzas.
At the left are the ten most frequent words in the poem by descending order of occurence.
This piece showcase one of those ten words.
Above is an area that directly represents the frequency of that word in each canto.
Each canto has a corresponding list of the ten most frequent words in that canto sorted by descending order of occurrence.
The length of the vertical lines for each canto represents its extension in number of verses.
The counted words were filtered before presentation by two factors — 1st No word with less than 10 occurrences in the whole poem would be taken into account. 2nd As you can imagine the most common words in Portuguese language weren’t considered like adverbs and pronouns. Some verbs and other words without a specific relevancy for the extrapolation of any concept, weren’t taken into account either.
As you should see, number ten was the magic number chosen for this composition. Each canto is composed by eight verses. The result was exported as pdf too – I’ll get this piece printed for sure!
For a matter of curiosity, here is a previous study that originated the final concept. This study displays each of the 50 most frequent words in Os Lusíadas along with the position of each occurrence in the text. Here are the first 21.
Almost about a month ago I made a small audio reactive composition that tries to attain visual richness through a simple concept like brownian motion. The piece was wrote with a particular soundtrack in mind — Kriespiel by Patrick Wolf. Having that in mind, although the composition reacts to any audio input, its feeling and timing might not be the appropriate ones considering other audio sources.
Due to the complexity of the piece, an offline rendering is required mainly because of the most prominent moments. The composition simply reacts to the energy of several sound frequencies. I used Processing’s sound library Minim. As this library doesn’t have any sort of synchronization mechanism towards Processing’s effective rendering frame rate, the audio was first pre-processed and later fed to another sketch that simply saves each frame of the composition. Audio and video were later mixed outside Processing.
Having said that you are free to scratch, reuse or even rip to pieces the code behind the composition. Please note that the code wasn’t thought for distribution and with that I state it’s buggy, isn’t documented and not very elegant. To use it you’ll have to really look at it. Please let me know about any forks of this composition.
A note about the distribution: you’ll find two sketches. The process_audio takes an audio.mp3 — not distributed — on the data folder and generates a fft.txt of the analyzed sound. Just let it run until de sound finishes. The brown3 sketch takes a fft.txt and starts the rendering! You have already an example fft.txt — A Boy and a Portrait from Yoko Kanno — doesn’t matter really.
Download brown.zip
Another small note. You could notice that the original Kriespiel video has some flickering. This wasn’t intentional, although it happens to suit very well. The problem — I did some operations that modified structurally (adds and removes) some data structures while at the same time I was drawing shapes corresponding to the elements of those data structures. The solution — always update first all your data structures and after you can draw the shapes! Pretty dumb straightforward.
This is mainly an experimentation with soft bodies using toxi’s verlet springs in Processing. The first idea was to visualize the greatest empires decline. Along with that came the idea of fluid and timeless boundaries, and thus some kind of soft bodies dissolution.
Visualizing empires decline from Pedro M Cruz on Vimeo.
Those are some screenshots displaying the springs in the system. In white we have the springs that form each shape’s skeleton. There are other more robust configurations but as the forces were minimized the shape kept it’s body like behavior. The collisions were implemented using the red springs — center to center connections that repulsed at a minimum distance.
The data refers to the evolution of the top 4 maritime empires of the XIX and XX centuries by extent. I chose the maritime empires because of their more abrupt and obtuse evolution as the visual emphasis is on their decline. The first idea to represent a territory independence was a mitosis like split — it’s harder to implement than it looks. Each shape tends to retain an area that’s directly proportional to the extent of the occupied territory on a specific year. The datasource is mostly our beloved wikipedia. The split of a territory is often the result of an extent process and it had to be visualized on a specific year. So I chose to pick the dates where it was perceived a de facto independence (e.g. the most of independence declarations prior to the new state’s recognition). Dominions of an empire, were considered part of that empire and thus not independent.
I don’t wanna call this small experiment of information visualization neither information art. Either way sounds too pretentious — as the visuals are not very sophisticated or elegant, and the way that the information is treated doesn’t enable the extraction of advanced knowledge. Although, it works very well as a ludic narrative. I ultimately found it very joyful. A direct interaction with the timeline could be a future plus!
The most interesting thing is that although not very extent, this data can be worked and displayed in several ways. More work on that, perhaps, later.