Thoughts on learning and work

Day 8 of 180 Days of Data Viz Learning #jfdi


I’m doing some form of data visualization learning for 180 days straight because I need to #JFDI.  See post explaining how and why I’m doing this.

Visualization Worked On or Created: 

N/A Today – Last Friday in LA before my move, spent it with my Mom and went out with friends

Decomposition of a Visualization:

75 mass shootings since Sandy Hook, in one map

  • What are the:
    • Variables (Data points, where they are, and how they’re represented):
      • Locations on map(shape), Description,
    • Data Types (Quantitative, Qualitative, Categorical, Continuous, etc.):
      • Quantitative, Categorical
    • Encodings (Shape, Color, Position): Position
  • What works well here?
    • Showing frequency and geographic spread
  • What does not work well and what would I improve?
    • Too much info hidden in metadata, you have to click each point for each story.  There should be another graphical form, perhaps in a filterable list, to look by location and victims and other data points.
  • What is the data source?  Do I see any problems with how it’s cited/used?
    • Stanford Geospatial Center
  • Any other comments about what I learned?
    • I think there’s a case of limiting factors in terms of tools that make this less effective than it could potentially.  The Google Map requires clicking in individual data points, and the card stack is not the best for consuming long-form information, where the meat of the analysis is in this piece.

Code Learning:

Udacity Problem Set One Finished – On section 2a Design Principles

Three Takeaways:

  • When you want to show how something has changed over time, use a line chart.
  • When you want to show how something is distributed, use a histogram.
  • When you want to display summary information, use a table.

Aligned Left Tutorials back on track

I decided to do this overall tutorial now because Udacity really gave me a good and deep fundamental understanding of how everything worked together, especially as Javascript and Front-end newbie.  I had started the full book Interactive Data Visualization and stopped when it became clear I didn’t understand as big of a picture as I wanted.  Part of what I’ve learned throughout this process and grad school is how to optimize learning resources, it’s definitely fail fast and learn from what’s most efficient for you right now, even if it takes some experimentation and tries on different tutorials and pacing.  I got to a point in Treehouse where I think I’d like more practice on basics and from the Udacity course, so I’ll revisited that in a bit.

Three Takeaways

  • View Source” shows you the original HTML source of the page, the web inspector shows you the current state of the DOM. This is useful because your code will modify DOM elements dynamically. In the web inspector, you can watch elements as they change. You’ll also use the JavaScript console for debugging.
  • Many, but not all, D3 methods return a selection (or, really, reference to a selection), which enables this handy technique of method chaining. Typically, a method returns a reference to the element that it just acted upon, but not always.
  • When chaining methods, order matters. The output type of one method has to match the input type expected by the next method in the chain. If adjacent inputs and outputs are mismatched, the hand-off will function more like a dropped baton in a middle-school relay race.

Reading and Learning Data Visualization Theoretically/Critically:

How NOT to Lie with Visualization. Bernice E. Rogowitz, Lloyd A. Treinish from 7 Classic Foundational Vis Papers

Three Takeaways:

  • “Without guidance about the physical or psychophysical properties of color, or about which colormaps are most appropriate for which types of data, the user is at a loss, even if the system provides a colormap editor or a library of pre- computed colormaps.”
  • “For nominal data, objects should be distinguishably different, but since the data themselves are not ordered, there should be no perceptual ordering in the representation. For ordinal data, objects should be perceptually discriminable, but the ordering of the objects should be apparent in the representation. In interval data, equal steps in data value should appear as steps of equal perceived magnitude in the representation. In ratio data, values increase and decrease monotonically about a true zero or other threshold, which should be preserved in the data representation.”
  • One important application of scientific visualization is to represent the magnitude of a variable at every spatial position. In many cases, the interpretation of the data depends on having the visual picture accurately represent the structure in the data. In order to accurately represent interval data, for example, the visual dimension chosen should appear continuous to the user. Candidate colormaps which preserve the monotonic relationship between data values and perceived magnitude can be drawn from psychophysical scaling experiments.

Eenie, Meenie, Minie, Moe: Selecting the Right Graph for Your Message

Three Takeaways:

  • Categorical data tells us what and quantitative data tells us how much. Quantitative data without related categorical data is useless.
  • A nominal scale has no intrinsic order. Is there an intrinsic sequence to these departments: sales, operations, finance, human resources, and IT? You might list the departments of your company in a particular sequence based on convention, but the list has no particular inherent order. The term nominal—”in name only”—suggests a set of items with different names, but no particular relationship to one another.
  • An ordinal scale has an intrinsic order. The terms “ordinal” and “order” are semantic siblings. Perhaps the simplest categorical scale of the ordinal type is first, second, third, and so on.