CREATION

How to create a data visualization?

Design rules, visual encodings, visualization types, tools, animation and interactivity options.

How to interpret data visualizations?

Pattern discovery, trend analysis, outlier handeling, storytelling

INTERPRETATION

How to evaluate a data visualization?

Human perception rules, user experience, showing the truth, chartjunk, visual space efficiency

EVALUATION

The purpose of data visualization are insights not pictures

Let my dataset change your mindset

It's not what you look at that matters, it's what you see.

The purpose of data visualizations is to enlighten people.

Not to entertain them, not to sell them products, services, or ideas, but to inform them.

It’s as simple and as complicated as that

Above all else show the data.

BIG DATA

Big Data Viz

Now, through ADV [Advanced Data Visualization], potential exists for nontraditional and more visually rich approaches, especially in regard to more complex (i.e., thousands of dimensions or attributes) or larger (i.e., billions of rows) data sets, to reveal insights not possible through conventional means.

EVENT DETECTION 

A radial plot visualization of 23,581 photos taken during 24 hours in Brooklyn area during hurricane Sandy (2012).

The photos are organized by time (angle) and hue (distance from the center).

Note the line marking a change in the number of photos and their brightness, corresponding to the moment of the power outage in the area.

This sudden and dramatic visual change (compared to normal day) reflects well the intensity of the human experience during the event.

PREDICTION

Prediction skill

How well the prediction model performed on historic data in that region is expressed through opacity.

Regions with higher skill values are more opaque, regions with lower values more transparent.

Only lines that represent areas with skillful forecasts are visible at all.

Summarises the mean wind speed observations in the selected geographic region over the last 30 years, in the respective season. Each observation is color coded and split into three categories of equal size (terciles) that indicate if that year had low(blue), medium(grey) or high(yellow) wind speeds.

Observation plot

Presents the predicted values of the 51 ensemble members obtained by the RESILIENCE prototype. These values comprise the calibrated probabilistic prediction for the selected region. The ensemble predictions are displayed as a cone of rays emanating from the typical (median) value of the historic data.

Prediction plot

SIMULATION

STORYTELLING

ALGORITHM

ALGORITHM VIZ

To visualize an algorithm, we don’t merely fit data to a chart; there is no primary dataset. Instead there are logical rules that describe behavior. But algorithms are also a reminder that visualization is more than a tool for finding patterns in data. Visualization leverages the human visual system to augment human intellect.

DATA ART

INFORMATION/DATA ART

Information art has a long history as visualization of qualitative and quantitative data forms a foundation in science, technology, and governance. Information design and informational graphics, which has existed before computing and the Internet, are closely connected with this new emergent art movement

A series of posters designed to resemble constellation maps but instead of being based on real stars, the shapes are based on first sentences from chapters of short classic stories.

Literary Constellations

When starting on a new data visualization project, I usually ask for two things from my clients: an objective and a data set.

To me, it is essential to understand what we want to achieve on a strategic level, but also which inspirations we can draw from investigating the data. Sometimes, the data can suggest ideas and insights you could not have come up with on your own.

Too much chartjunk

Bad choice of colours

Missing focus

The grey background, the grid, the shapes on the lines, the frame around the legend, the axis labeling

In this example colours have the function to distinguish, therefore choosing similar colours makes no sense.

Including many lines in one chart without the option to help the user focus on one and compare to the others, interpretation is difficult.  

HIGHLIGHTING

Highlighting the lines to provide data context while focusing on select series, which allows to explain the feature of one particular group compared to the others. Make it appear different, and give it a proper annotation.

In this example:

Trough the size of the life: highlighting the average, extreme values, discovered groups of lines representing a certain pattern,

FILTERING

Filtering simply eliminates elements, it is very straightforward for users to understand and compute. 

The challenge comes in designing a vis system where filtering can be used to effectively explore a dataset.

In this example:

The user types in a search box the terms which wants to be compared.

Is a series of similar graphs using the same scale and axes, allowing them to be easily compared. Comparing two views that are simultaneously visible is relatively easy, because we can move our eyes back and forth between them to compare their states. In contrast, for a changing view, comparing its current state to its previous state requires users to consult their working memory, a scarce internal resource.

In this example:

The level of immigration/emigration it layered by country. The ordering helps the user interpreting the results and detect extreme values and common patterns.

SMALL MULTIPLE

LINKED HIGHLIGHTING

Interactivity unleashes the full power of linked views., the most common forms of linking is linked highlighting, where items that are interactively selected in one view are immediately highlighted in all other views using in the same highlight color. whether it is distributed differently..

In this example:

The category state is highlighted across the different KPIs: population, assaults, robberies and murders.

LINKED HIGHLIGHTING & FILTERING

In this example:

The data is represented as a small multiple using the category "theme", highlighting the current year compared to the rest of the years and filter on different years via a button applied to the complete visualization.

Solutions

Change symbol size > trying to increase white space

Use transparency > when more symbols appear on top of each other leads to less transparency

Reduce data > break up the population into sub samples and show them via small multiples

Aggregate data > into bins

WTF

Saturation

Don't over do it !!!

Use it to guide the viewer, tell the story, change the mood or draw attention to something

Show the data

Tell the truth

Help to think about information rather the design

Encourage the eye to compare the data

Make large datasets coherent