Started getting a thought process down on how to approach building a new visual -> eg starting at x y coordinate and moving to changing variables
Look in dev console for missing pieces and change code blocks bit by bit, there’s a tendency with D3 to want to make a lot of changes at once – at least on my end, but don’t forget coding fundamentals when approaching these problems
The console is your friend in testing out changes to scales and domains
John Snow visual review – > first data viz, bar chart pattern, looking at data points that refute hypothesis
Look for exceptions in visuals to verify methodology
Symbiosis between words and visuals -> annotations crucial to narratives -> prevent from people from drawing wrong conclusions
Particularly needed for multi-section visuals
NYT and Pew do really good job by putting in text and story
Sequencing info “The upper layer” to giving a broad view and context to “layering down” and give multiple ways of exploring maps
“Splitting-a-sentence” strategy – > write headers as if they were a very long sentence that describes entire process and sequence and find natural breaks and transform each one of portions into headers of each section
Knight Journalism Course for the Americas: Data Visualization and Infographics with D3!
Three Takeways from Module 4 Scaling and Axes (covering topics I thought I knew well but apparently do not so just ret-typing for remembering).
The proper sequence of methods to use when binding data and creating new elements from data: .selectAll()data()(enter()append()
“Dom is current state of memory of a web page that can different from page’s source and CSS”
Ordinal Scales have a discrete domain, such as set of names or categories -> values must be coercible to a string, and the stringified version of the domain value uniquely identifies with the corresponding range value
+ symbol means “take this value and use it as a number” after a return statement