Everyday, people invent more and more ways to see the world differently. They create new relationships between various bits of information to invent newer perspectives. Because there's so many different ways to see the world, there will come a point when just reading about it won't do: there will need to be visualization.
In a visualization, data is removed from its static, one-dimensional existence and placed in a space that is many-dimensional and dynamic. In such an environment, the data's virtual position also has meaning, like an electron's position around a nucleus tells us something about its energy.
This ability to add more meaning to information based on how we perceive it makes all the difference because this brings numbers closer to reality and lets us engage with them more directly. Instead of requiring mathematics and statistics to mediate between us and what the data mean, visualizations let us use our eyes and eyes to decipher the real meaning.
My search for a good tool with which to create visualizations has led me to discover a lot of options. At the same time, understanding how a visualization has been created tells me something about how its creator was perceiving the data. Sometimes, there are no omnipresent creators: a software need only be built that, in turn, can be used to create other visualizations, and voila! We're suddenly using machines to seek out curious patterns that haven't yet hit upon.
Here's a list of some of the tools I use to visualize the statistical information I access. Mostly, I use them to illustrate a point - in class or in my posts. However good writing is praised, creating dimensionality is tedious with pen and paper. At the cost of discouraging long didactic essays, I'd say use tech. to make your point. After all, each one is centered on good and purposeful communication.
Ngram Viewer is one of the more popular services of Google, at least amongst the ones that weren't widely publicised. An n-gram, in computational linguistics, is a contiguous sequence of 'n' items. For instance, a one-gram is one word, a two-gram is two words that are counted as one phrase, a ten-gram is a set of ten words taken together, and so on. Using the millions of books that Google has parsed through to create its database - or corpus - Ngram Viewer plots the frequency with which an n-gram appears in all books over time. The idea is that the more an n-gram appears in books, the more widely it was spoken about in those years.
Fusion Tables is a service by Google that interpretes numbers in different ways. The user has to upload a database - which might comprise one table or hundreds - and then pick one of many visualization options. For instance, I could upload a table that contains information on where how much pollution is. Next, if I select the option to plot the table on a map, Fusion Tables immediately does so. The service is useful in generating newer perspectives as Ngram Viewer, too, can and also to make a point crisply, efficiently.
Wolfram Mathematica is a computation engine, i.e. it performs a wide variety of computations - mathematical, statistical and logical - and then chooses the best way to display the information. This clear understanding of how many dimensions information has is what makes this possible, and I suspect it's the heartless mathematical treatment at its core that's responsible. WM is not for people who don't know what they're looking for. It's for those who know their way around, but are just looking for the best ways to express what they're saying.
Microsoft Robotics Developer Studio
OK, I discovered the visualizing potential of MRDS by accident. Because it uses a visual programming language with flow charts and discrete processes to simulate robotic behaviour, it becomes easier to understand what's going on inside an automaton's programmatic head. Using the same principles, an interactive visualization can be simulated, with the user controlling the robot that's being moved around. It's kinda abstract, I know, but making a higher level of interaction possible, we're only making the visualization more engaging, memorable and understandable.
Visualizations are not always about numbers; they're about spatial geometries, too, and when it comes to them, parameters like distance, time and speed are more perceptible than colours, gradients, and shapes. And what makes MRDS more capable in this context is its ability to simulate, which should be exploited.
Now, d3.js is a real beauty. Before I say anything, I'll say this: when it comes to programming, I'm the type who gets quite excited about it because I love algorithms, but on the downside, I'm not so good with the syntax. That is, I can get what steps a program takes to accomplish something but I can't get how really goes about doing it. On that note: d3.js is really intuitive semantically and demands only a few hours' practice to make sense syntactially, too. And because it's a library focused on linking documents with Document Object Models (DOM), it's the ideal place to begin for people wanting to create their visualization up from scratch.
No comments:
Post a Comment