IS498 mid term
Visual analytics
analytical-style visualisation work, such as dashboards, that serve the role of operational decision support systems or provide instruments of business intelligence.
Exhibitory visualizations
are characterized by allowing the user to interpret the data on their own. The users have to rely on their own capacity to perceive and translate the features of a visualization.
Simple,
matter that is inherently easy to understand
Mindsets:
thinking, doing, making
Data journalism
adaption of data visualisation but with unquestionably deeper roots in the responsibilities of the reporter/journalist
not always needed in exploratory data analysis
Not all visualisation challenges will involve much EDA and not all visualisation projects will give you space to do much EDA.
Data science
gathering, handling, analysing and presenting data
supplied data
getting it from somebody else
data
names and amounts. It is groupings, descriptions and measurements. It is dates and locations
Infographics
traditionally created for print consumption contain charts (visualisation elements) but may also include illustrations, photo-imagery, diagrams and text produced for static output
The Three Principles of Good Visualisation Design
trustworthy, accessible, elegant
Feeling tone
visual quality that embodies a feeling tone of voice might be described using adjectives like emotive, seductive, figurative, big-picture, fun and dramatic.
perceiving
what do I see? What chart is being used? What items of data do the marks represent? What value associations do the attributes represent? What range of values are displayed? Are the data and its representation trustworthy?
Information visualisation
work that is primarily concerned with visualising abstract data structures such as trees or graphs (networks) as well as other qualitative data (therefore focusing more on relationships rather than quantities)
cleaning data for transformation
'find and replace' (or remove) operation. Other tasks might be much more intricate, requiring manual intervention, often in combination with inspection features like sorting or filtering, to find, isolate and modify any problem values
vision
'the ability to think about or plan the future with imagination or wisdom' what is its purpose?
why important to follow design process
-reduce randomness of approach -every project different -adaptability -protect experimentation -1st occassion, not last -time mgmt -mindsets -documenting -communication -attention to detail -kill your darlings -learn
Storytelling
Charts showing trends or activities over a temporal plane or maps portraying spatial relationships offer displays that are most consistent with the idea of a story
What are activities to work with data described in Chapter 4? List and briefly define.
Data Acquisition - This is basically where the data set comes from, data sources such as web scraping, supplied, system download, APIs. Data Examination - Familiarizing yourself with the data that has been collected, the data types, what condition it is in, the size and range. Data Transformation - This is preparing the data for analysis and charting, cleaning, creating, consolidating. Data Exploration - discovering insights from the data using exploratory analysis and research techniques.
What are the five distinct design layers that make up the anatomy of any visualization solution? These are covered in Part C of the textbook.
Data representation Interactivity Annotation Color Composition
'Kill your darlings'
Even though you have invested heavily in time and emotional energy, do not be stubborn. When something is not working, learn to kill it.
consolidating data for transformation
Expand: This is where you want to broaden the values of data you have to work with. Append: This might occur if your original dataset is no longer representative of the most up-to-date state and newer data items are available for you to access
Interval example
Forecasted temperatures in Celsius. The body mass index for measuring obesity.
What are the three main elements/goals of formulating your brief for visualization design according to Kirk?
Initiating, defining, and planning requirements of the project.
Describe one way to create new data that will not compromise the integrity of the data for your visualization. These are described in the "Creating" section of the 4.3 subheading of the textbook. The techniques described here are all valid methods of data transformation. Select one and describe it in your own words.
One way to "create" new data as explained in the book would be using the "start date" and "end date" values to calculate the number of days that something took. This is "creating" data because its not explicitly coming from the dataset but it is something easily derived from it.
Mark
Points, lines or shapes that represent items of data
Ordinal example
Ranks of police officers. General temperature observations from "very hot" to "very cold".
Textual example
Responses to "Any other comments?" in a questionnaire. The abstract for an academic research article.
when to simplify vs clarify
Simplify when your audience do not have the knowledge or capacity to handle a complicated subject and do not need to acquire deep understanding about it Clarify when your audience do not have the knowledge but do have the capacity to handle a complicated subject, with assistance.
List the four stages of the data visualization design process. Define each briefly.
Stage 1 - Formulating your brief, this involves planning, defining, and initiating your project. Stage 2 - Working with data, gathering, handling, and preparing your data. Stage 3 - Establishing your editorial thinking, defining what your will show your audience Stage 4 - Developing the design solution, the visual manifestation of the preparatory work you have conducted
Who are the people you need to research when formulating your brief for visualization? Name each group and describe their role in visualization design briefly.
Stakeholders - The people that have a stake in the visualization, whether they are the ones that are funding the development or subject-matter expert. Audience - The target audience for the visualization, the people you wish to reach by developing it. Visualizer (s) - Those that are involved with the project to help bring it to life, these can be different people or someone wearing/doing multiple "hats"/roles.
Data exploration involves the use of both statistical and visual techniques. (t/f)
T
Representing the online sales channel data from Chapter one in a visual chart IS BETTER than representing it in a table. (t/f)
T
Ratio example
The ages of survey participants in years. Forecasted amounts of rainfall in millimeters.
Match the design guidelines with their elements. Good visualization design is accessible and usable
The design of the representation and the presentation is suitably understandable The portrayal of the data and the subject is relevant
Match the design guidelines with their elements. Good visualization design is trustworthy and reliable
The handling of the data is reasonable and faithful to the subject The representation and presentation design have integrity
Match the design guidelines with their elements. Good visualization design is elegant and aesthetic
The representation and presentation design is appealing
how representation and presentation connected example
The selection of a chart type inherently triggers a need to think about the space and place it will occupy on your screen or page
comprehending
The viewers now consider what the interpretations mean to themselves What has been learnt? Has it reinforced or challenged existing knowledge? Has it been enlightened with new knowledge? What feelings have been stirred? Has the experience had an impact emotionally? What does one do with this understanding? Is it just knowledge acquired or something to inspire action, such as making a decision or motivating a change in behaviour?
The first occasion, not the last
There are some recommended habits that are applicable to all stages in this process, relevant to novices or experienced visualisers alike, as follows.
What are the four constraints you need to explore for each visualization project? List each and briefly define.
Timescales - How much time to develop the project. There is a starting time and an ending deadline, that can be set by stakeholders or yourself. The starting time can be dependent on when you can finalize the dataset. Pressures - External pressures such as budget/costing, staffing, politics of the subject, cultural sensitivities, even environmental. The politics of the subject can be at odds with the integrity of the project. Design - Style guidelines for specific colors, typeface, fonts, and logos. There can also be layout and size restrictions depending on the medium the visualization is being produced for. Technological - The tools and libraries used to create the visualization but also the way that the audience will consume the visualization should be taken into account.
What is data visualization in Kirk's terms?
Visual representation and presentation of data to facilitate understanding.
Attributes
Visual variations of marks to represent values associated with the marks, properties such as size, color, position
Interpreting
What does what you have seen mean, given the subject? What features - shapes, patterns, differences or connections - are interesting? What features are expected or unexpected? What features are important given the subject?
Documenting:
Whether using pen and paper, or a tool like Word or Google Docs, note-taking is a useful habit to develop. It helps you document important details
You do not visualize because you have ----. You visualize because want to---- something about the ----.
You do not visualize because you have data. You visualize because want to understand something about the subject.
research in exploratory data analysis
You need to explore the places (books, websites) and consult the people (experts, colleagues) to give you the best chance of getting accurate answers to the questions you have
Exploratory visualizations
are focused on helping the viewers or the users discover and form their own interpretations. Almost universally, these types of tools are interactive to allow the user to explore the visualization.
Primary collection
collect observations or capture measurements about bespoke phenomena specific to your needs
presentation
design choices such as the possible application of interactivity, features of annotation, all matters around colour usage, and the composition of the work
Dashboard
displaying multiple visualisations and statistical information.
Scientific visualisation
drawing out the scientific methods for analysing and reasoning about data for conceiving highly complex and multivariate datasets specifically concerning matters with a scientific bent
Time management:
each project introduces its own profile of demands, so always find time before you set off to estimate where your likely commitments will be most required.
APIs
enable people to access streams of data.
Attention to detail:
errors found in your work can be damaging and will certainly undermine your audience's trust
Which of these phases make up the "hidden thinking" of data visualization according to Andy Kirk? These are covered in Part B of the book. Select multiple answers. Establishing your editorial thinking Working with data Developing the design solution Formulating your brief
everything except #3
creating data for transformation
expand your data to form new calculations and derive new groupings or any other mathematical treatments
data size concerns (4)
frequency distribution, central tendency, frequency counts, measurements of spread
reasoning in exploratory data analysis
help reduce the size of this prospect Deductive reasoning is targeted: You follow a specific curiosity or hypothesis, framed by subject knowledge, and interrogate the data Inductive reasoning is much more open in nature: You 'play around' with the data, based on your sense or instinct about what might be of interest, and wait and see what emerges
Communication
listening to stakeholders and to your audience: what do they want, what do they expect, what ideas do they have? In particular, what knowledge do they have about your subject?
Learn:
looking back over your work, examining the output and evaluating your approach.
What is visual presentation in Kirk's terms (Chapter 1) and what are elements of it?
making decisions about how to portray the data visually so that the subject understanding it offers can be made accessible to the audience. Basically it is the charts and which one shows the correct element of the data that you are trying to explain/most relevant to the data.
visual representation
making decisions about how you are going to portray your data visually so subject understanding made accessible to your audience.
Data foraging
manually sourcing relatively small amounts of disparate or dispersed data values
Instinct of analyst in exploratory data analysis
needs to possess the capacity to recognise and pursue the scent of enquiry. natural inquisitiveness and the sense to know what approaches (statistical or visual) to employ and when.
Every project is different
no two instances of that report will involve the exact same context
nothings in exploratory data analysis
nothing usually still means something. Reaching a dead end or going down blind alleys can be helpful because they help you eliminate dimensions of possible analysis.
Explanatory visualizations
offer an experience characterized by the visualizer taking responsibility to present important observations and interpretations to help the viewer more quickly assimilate the meaning of what is presented.
Protect experimentation:
one must still seek out opportunities - in the right circumstances - for imagination to blossom
Reading tone:
optimising the precision and efficiency of perceiving the represented data.
Give a brief summary of the example the book gives for the benefits of visualization (contrast table vs. graph).
organized raw data, you can only answer/interpret a basic understanding of trends. graphs can highlight the full capabilities and findings of the data, which can be enhanced with design elements such as color, interactivity, annotation(s), and composition.
understanding (data vis definition)
perceiving, interpreting and comprehending
Information design
presentation of information design of many different forms of visual communication, particularly those with an instructional or functional slant
Data art
pursuing a form of self-expression or aesthetic exhibition using data as the paint and algorithms as the brush
tones
reading, feeling
facilitating
realistically the most a visualiser can do There are times when the onus is on us, and other times when the onus is on the viewer.
Adaptability
respond to revised requirements, additional data that emerges, or a shift in creative direction
Deliverables
setting, medium, quantity, frequency
Reducing the randomness of your approach
shapes your entry and closing points. How do you start a process? How do you know when you have finished?
Web scraping
special programs to extract structured and unstructured items of data published on web pages
Complicated
subject knowledge or a skill that is typically intricately technical, probably unique and difficult to understand.
Complex
systems or contexts that have no perfect conclusion or even no end state.
system download
t let interested users construct detailed queries and download structured data customised to their need