(in reverse order)
20211123
Tica led us through an Airtable-to-Unity tutorial, showing us how to create a virtual gallery using data from an Airtable.
20211116
Zane did a D3 tutorial!
20211109
@marlon will do a little blender tutorial!
This document is going to walk you through a basic D3 example. You're welcome to just read the document, but I definitely recommend opening the basic CodePen here and coding along as you go. Playing around with the code yourself, seeing what changing this or that does to the final result, is one of the best ways to learn.
CodePen
Also, the screenshot text appears small in this document, I definitely recommend zooming in on the page, hitting Ctrl + + a few times until you can see everything clearly. No need to ruin your eyes. :)
The Modern Webpage (in a paragraph)
Webpages have three distinct components, generally. The HTML document (a file that ends in .html) is the file that houses the content of the webpage - the name, the header, the paragraphs of content, the images. The CSS file (ends in .css) is the style sheet. It tells the HTML file what fonts to use for the headers, where to place the paragraphs on the screen (left, right, center), what size the fonts should be, etc. And then the Javascript files (end in .js) are the files that make things (like graphs and charts), the results of which are delivered to the HTML document to be included in the webpage. Coding an interactive data visualization for a webpage involes making these three types of files and coding them all to get along.
The HTML Document
Zane Wolf changed 4 years agoView mode Like Bookmark
Enoding data is the process of converting or translating raw data into understandable (and hopefully rememberable) images or symbols. But before we can start encoding, it's important that we go over some terminology.
Data Types
First, what type of data do you have? Some types of encoding methods work better for some data versus others, and looking at your data, understanding what it is comprised of, should always be the first step. There are several different types of data.
Qualitative/Categorical
Categorical data are often data that can't be counted or measured, but can be divided into categories. Geographical Location, Color, Age Groups, and Months are all examples of categorical data. Within categorical data, there are two subtypes: nominal and ordinal. The difference between the two is whether that is an inherent order to the data. Geographical location....does alphabetical make sense? Maybe, but that's a choice of ordering that you could impose upon the data. Color also doesn't really have an order to it. But Age Group and Months do have an inherent order to them. Sizing (small/medium/larger) and Grades (A+/A/A-/B+/...) are both ordinal.
Quantitative/Numerical
Zane Wolf changed 4 years agoView mode Like Bookmark
Codepen
Planning Notes
Make a bubble chart, bar chart, or line graph using the Volcano Data set
bubble chart where x axis = time, y-axis = elevation, size = VEI?
scales/axes, could include a filter too!
could be fun to boot the final project from this to blender?? 3D bubble chart??
Zane Wolf changed 4 years agoView mode Like Bookmark
The last step of every project is communicating your results to other people. Whether it is a research project or a class you collected performance data for, you'll need to communicate what you learned, and often this involves showing the raw or processed data.
Data visualization, the display and meaningful communication of data, has often been overlooked or underestimated as the last and inconsequential step of a project. Surely you've seen impossible pie charts or misleading bar charts on news channels. There are many graphical principles and statistical best practices that should be observed when designing charts and figures, lest the visualization become illegible at best and actively dishonest at worst. The collection of these best practices and principles has become substantial enough that data visualization is now considered its own field of study. People can even have careers as "data visualizers" (among many other possible terms like "information designer").
Accompanying, or perhaps even facilitating this boom in data visualization as a field, has been a boom in data visualization tools. The purpose of this blog is to survey the different types of tools and the factors that you should consider when deciding which tool is right for you and your data.
Factors
Before settling on one specific dataviz tool, it's important to ask yourself a couple of questions. The answers to these questions will determine what set of tools (even which specific tool) would be the best one for you to use.
1) Do you know, or want to learn, how to code?
Zane Wolf changed 4 years agoView mode Like Bookmark
Link to Repo
Overview
These are example datasets that are open source, cleaned, and free to use for whatever project you would like to use them for.
LL Basics: Datasets
Overview
These are example datasets that are open source, cleaned, and free to use for whatever project you would like to use them for.
Names in Lyrics
Zane Wolf changed 4 years agoView mode Like Bookmark
Data link:
SEAS Research Interests https://github.com/SEASVis/SEASVis.github.io/blob/main/data/Final%20Visualization%20Data_People.csv
SEAS affiliated schools & centers
https://github.com/SEASVis/SEASVis.github.io/blob/main/data/Visualization%20Data_Centers.csv
Zane
Sketch
Tica
Data Type
Zane Wolf changed 4 years agoView mode Like Bookmark
from Tica's awesome November 2nd challenge!
One idea = having a kind of "fingerprint" for each center showing the distribution of faculty by rank. So that at a glance you get a sense of whether the Center is entirely comprised by junior faculty, say. That's what I'm trying to do below anyway.
I was also wondering whether we could focus viewers attention a bit more when giving the map of a given center's faculty and those faculty members' research areas. We could still start with a view that has EVERYTHING visible, but a modal element could pop up (or on large screens maybe a 60% width huge sidebar say) that really pulls focus to the "visual story" of a specific center.
So in the below image I'm trying to imagine a template for a "single-center-story" . . . presumably it could also involve a relatively big chunk of text we get from the center that tells the story of what they're up to (so that we enrich the micro-narrative we get from each new click with some prose that orients us a bit)
Finally there's an idea for a streamgraph of research areas to show them over some X value. Time would be ideal if we had snapshots of the data going back decades. But if we mainly have the present then, as Claire suggested, faculty rank could serve as a proxy for time?
Marlon Kuzmick changed 4 years agoView mode Like Bookmark
One initial thing we thought could be useful as a "launch" activity would be a workshop where we teach our Learning Lab Undergraduate fellows (and staff) some of what you know, then give them the task of mapping things related to our studio:
a simple map of how to get to the studio
a simple map OF the studio and what's in it
a more technical map of the signal-routing in our studio (we have of the order of 200-300 cables running throughout the space)
but then we are also starting to encounter a number of courses that are interested in the "mapping-as-storytelling" idea...
map as game-board
map as interpretive illustration of a fictional world
k-gilligan changed 4 years agoView mode Like Bookmark
tutorial in use:
https://dataviztoday.com/mapbox
challenges for future data viz endeavors/tutorials
downloading materials/getting log-in sorted out in advance
getting everyone on board about changes of plans
potential solutions for future data viz endeavors:
k-gilligan changed 4 years agoView mode Like Bookmark
Details
Date: 10/12/21
Time: 4 pm
Location: the Fireplace Room
Attendees: ~15
Group: Cluster 1
Timeline of Events
First, everyone had to get situated with Mapbox.
k-gilligan changed 4 years agoView mode Like Bookmark
Prep
Download the data
Make an account with mapbox studio
Pick an image that you would like to use as a color palette
For example, I used my desktop image .
Hovering over the circles tells you whether this will be color of Land, Greenspace, Water, Roads, or Labels.
Optional: Bring/find a second dataset (latitute/longitude must be included) to work with after completing the first tutorial.
k-gilligan changed 4 years agoView mode Like Bookmark
Okay, so this cluster may need to exist more like a constellation of unique stars, lighting up a galaxy of interests around data visualization, mapping, augmented reality, and more.
the Slack Channels
#ll-map-lab
#ll-data-vis-lab
#ll-reality-lab
the People
k-gilligan changed 4 years agoBook mode Like Bookmark