Find examples of data presentations that you consider to be particularly insightful and/or powerful. Discuss their strengths and weaknesses.
Find examples of data presentations that you consider to be particularly misleading and/or useless. Discuss their strengths and weaknesses.
How do you think new technologies (e.g. virtual or augmented reality, 3D-printing, wearable computing) will influence data presentations?
Consider the following datasets:
Create a data dictionary for this dataset. Establish a list of variables that you think are crucial to a good understanding of the dataset. Justify your choices.
Create (at least) 5 bivariate/univariate visualizations that can help you understand the dataset.
Produce (at least) 3 “definitive” visualizations for the dataset. Use the principles discussed in class (including documentation, legends, annotations, Multiple I’s, etc.). Emphasis should be placed on content AND on presentation (suggestions: consider creating a reasonably high number of charts using a random selection of a random number of variables in order to minimize the odds of missing out on useful information).
Repeat the previous question with any dataset of your liking.
Identify a scenario for which a dashboard could prove useful. Determine specific questions that the dashboard could help answer or insights that it could provide. Identify data sources and data elements that could be fed into your dashboard. Design a display (with pen and paper) with mock charts. What are the strengths and limitations of your dashboard? Is it functional? Elegant?
At what point in the data science workflow do you think that visualizations of this nature could be useful?
What are the ways in which observations could be anomalous? Have you found any such anomalies? Do you have explanations for them? In particular, consider the case of South Africa in 2012, which appears to be a clear outlier. Follow the path of the South African bubble from 1975 to 2020, in relation to the general pattern. Does the apartheid/income inequity explanation suggested in the text still make sense?
Pick 2+ “definitive” visualizations (methods, variables, etc.) other than the default configuration. What are some important insights?
How would you describe the insights of step 3 without resorting to visual vocabulary?
Can you think of ways in which the data of interest to you in your day-to-day activities could benefit from the same treatment? What situations could you explore in such a scenario? How would that help your team better understand the system under consideration?
Consider the following Australian population figures, by state (in 1000s):
Graph the New South Wales (NSW) population with all defaults using
plot(). Redo the graph by adding a title, a line to connect the points, and some colour.
Compare the population of New South Wales (NSW) and the Australian Capital Territory (ACT) by using the functions
lines(), then add a legend to appropriately display your graph.
Use a bar chart to graph the population of Queensland (QLD), add an appropriate title to your graph, and display the years from 1917 to 2017 on the appropriate bars.
Create a light blue histogram for the population of South Australia (SA).