Module 3

[crangelsmith]

Finalising content for Module 3

[review-notebook-app[bot]]

Check out this pull request on 

See visual diffs & provide feedback on Jupyter Notebooks.

---

Powered by ReviewNB

[nbarlowATI]

Very nice! A few comments on the first couple of sections (other sections coming later):

Figures gone wrong

• Example 2 notes: "there is clearly two axes" => "there are clearly two axes".
• Also Example 2 notes: "Plotted these two data sources here" => "Plotting these two data sources together here"
• Example 4 notes: "Failrue to normalise" => "Failure to normalise".
• Also Example 4 notes: I don't think we mean "state population" - the plots are higher resolution than that... I think we should just say "population", and "vary across states" => "vary across the map".
• Example 5 notes: Not sure what I'm suggesting to change the text to, but I think the right hand plot comparing average incomes between USA and Mexico doesn't have the axis exaggerating the effect, but actually the opposite - if it were a linear y-axis, the difference would look even bigger!
• Also Example 5, while I agree that the kernel density estimates give a better understanding, they are not showing exactly the same data as the plots of the averages, in that they don't show the time variation (at least the static versions don't - I guess GapMinder normally have these as animations...)
• Examples 9 and 10: I feel these plots are a different category to the earlier examples - they are not exactly misleading, just have some bad aesthetic choices, which seems like a less serious problem... Maybe this is fine, just something for the teacher to mention.

Rules of the data visualisation game

• "underlined by the data" => "underpinned by the data" perhaps?
• "that need decided upon" => "that need to be decided upon"
• Colour scales section: "increased cases for wealthier country" => "increased cases for wealthier countries"
• "Continuous Colour" section: "The right-most figure gives an adjusted figure." This sentence is quite clunky, but also, the figures below are one underneath the other - it's not clear what the "right-most figure" refers to.
• "Colour to highlight" section - several uses of "color" rather than "colour".
• "Central role for colour" section: "full on nuance" =>" full of nuance", and "thorugh" => "through", and "use colour as" => "use of colour as".
• "Aggregation" section: "miscommunication trends " => "miscommunicating trends"
• Also in the "Aggregation" section, there is an error in the code cell: NameError: name 'df_uk_cases_1y' is not defined - it doesn't look like this dataframe is created anywhere.

[nbarlowATI]

Some more comments on the later sections of Module 3:

Atlas of Visualisations

• In the introduction: "Rules of the Data Vizualisation Game" => "Rules of the Data Visualisation Game", and "high level" => "high-level"
• "Visualising amounts" section: "Bars are a good option" - I fully agree with this statement, but maybe we should specify "Bar plots are a good option" :)
• For the bar plots of vaccination by US state, the caption says that the right hand plot is better, as it makes it more obvious which states are doing better. However, I would argue that the left hand plot, alphabetically ordered, is more useful if someone wanted to look for a particular state.
• In the caption for the dot plots, "ordered by the amount size " => "ordered by amount"
• Caption for heatmaps: "neceseraly" => "necessarily".
• "Kernel density figures" - the figure titles say "Ban width" rather than "Bin width".
• "Pie charts" - looks like there is a missing figure.
• "Scatter plot" example - the figure title has a typo, "expentancy"=>"expectancy". Also, I'm not sure about showing the regression line here, without any explanation of it in the text - wouldn't it be better to have it as a pure scatter plot?
• "Slopegraphs" section, "slopgraph" => "slopegraph". Also, couldn't one make the same point for this as was made for example 2 of "Figures gone wrong" - that this is implying there was a linear trend between the two points?
• "Uncertainty" section: "Any insight has always associated an uncertainty" => "Any insight always has an uncertainty associated with it."
• Same section: "and visualising can be" => "and visualising it can be"
• "can result on the public" => "can result in the public", and "refuse to comply" => "refusing to comply"
• "shortcomings on this area has been exploited" => "shortcomings in this area have been exploited"
• "we have not develop content" => "we have note developed content"
• "he highly recommend" => "we highly recommend"

Storytelling with data visualisation

• "starting point of your figure’s story" - maybe "starting point of your visualisation's story" - there might be more than one figure?
• "relatioships shown" => "relationships shown"
• "expandable example here" needs doing :)
• "Hierarchy of the information" section - if I were teaching this section, it's not really clear what I was trying to get across, or what to take from the figures - are some of the combinations of titles/subtitles, font sizes, labels, better than others?
• "Emotion and context in storytelling" section: "In this course we do not have a clear answer for this" - maybe instead say "There is not a single definitive answer to this".
• "Context" section: "tens of thousands first-time" => "tens of thousands of first-time"
• "does not succeeds to communicate" => "does not succeed in communicating"

Visualisation for data exploration

I feel that this section should be renamed something like: "Walkthrough: visualisation for data exploration" - it's really not the same format as the previous sections.

• First code cell throws an error: FileNotFoundError: [Errno 2] No such file or directory: './data/UKDA-7724-csv/csv/eqls_2011.csv'
• "Eductation and Health" => "Education and Health"
• Section heading "MISSIGNESS" => "Missingness", then also for the next section "Dealing with Missigness" => "Dealing with Missingness".
• "we need to careful balance" => "we need to carefully balance"
• "missigness jargon" => "missingness jargon"
• "droping rows" => "dropping rows"
• "we would be loosing" => "we would be losing"

Hands-on session

I guess it's obvious from the previous sections/hands-ons, what the world in data dataset, and "df" are?

[nbarlowATI]

Longer term suggestions

(i.e. not for this iteration of the course)

• Something about interactive plots, and animations - these can give another dimension of information in a natural way. (Gapminder of course have some good examples.)
• Perhaps a list of some popular visualisation libraries/packages, and short description of each, along with a pointer to their documentation.