Winning Ways for your Visualization Plays by Mark Grundland
1 like1,254 views
The document discusses different visualization techniques throughout history and their impacts. It begins by describing cave paintings from 35,000 BC and then discusses the invention of word clouds in 1992 AD. It next talks about how John Snow's visualization of cholera deaths in 1854 London helped identify the water pump responsible for spreading the disease. The document concludes by examining the controversial "hockey stick graph" showing modern warming is unprecedented, and how visualization design choices can influence data interpretation.
![Communicating data effectively can
change what we do with our world
Hockey stick graph was arguably the most controversial
chart in science: our future depends on how we read it.
8!
w
http://www.ipcc.ch/pdf/climate-changes-2001/synthesis-syr/english/summary-policymakers.pdf!
line, a filtered annual curve suppressing
fluctuations below near decadal
time-scales). There are uncertainties in
the annual data (thin black whisker
bars represent the 95% confidence
range) due to data gaps, random
instrumental errors and uncertainties,
uncertainties in bias corrections in the
ocean surface temperature data and
also in adjustments for urbanisation over
the land. Over both the last 140 years
and 100 years, the best estimate is that
the global average surface temperature
has increased by 0.6 0.2°C.
(b) Additionally, the year by year (blue
curve) and 50 year average (black
curve) variations of the average surface
temperature of the Northern Hemisphere
for the past 1000 years have been
reconstructed from “proxy” data
calibrated against thermometer data (see
list of the main proxy data in the
diagram). The 95% confidence range in
the annual data is represented by the
grey region. These uncertainties increase
in more distant times and are always
much larger than in the instrumental
record due to the use of relatively sparse
proxy data. Nevertheless the rate and
duration of warming of the 20th century
has been much greater than in any of
the previous nine centuries. Similarly, it
is likely7
that the 1990s have been the
warmest decade and 1998 the warmest
year of the millennium.
[Based upon (a) Chapter 2, Figure 2.7c
and (b) Chapter 2, Figure 2.20]
3
1860 1880 1900 1920 1940 1960 1980 2000
Year
Departuresintemperature(°
fromthe1961to1990avera
Departuresintemperature(°C)
fromthe1961to1990average
(b) the past 1,000 years
NORTHERN HEMISPHERE
Data from thermometers (red) and from tree rings,
corals, ice cores and historical records (blue).
1000 1200 1400 1600 1800 2000
Year
−1.0
−0.5
0.0
0.5
Data from thermometers.
−0.8
−0.4
0.0
0.4](https://image.slidesharecdn.com/winningwaysforyourvisualizationplaysmarkgrundland-140328093347-phpapp01/85/Winning-Ways-for-your-Visualization-Plays-by-Mark-Grundland-8-320.jpg)
![Communicating data effectively can
change what we do with our world
Hockey stick graph was arguably the most controversial
chart in science: our future depends on how we read it.
9!
w http://www.ipcc.ch/pdf/climate-changes-2001/synthesis-syr/english/summary-policymakers.pdf!
http://www.amazon.com/dp/023115254X/ and http://www.amazon.com/dp/1906768358/!
line, a filtered annual curve suppressing
fluctuations below near decadal
time-scales). There are uncertainties in
the annual data (thin black whisker
bars represent the 95% confidence
range) due to data gaps, random
instrumental errors and uncertainties,
uncertainties in bias corrections in the
ocean surface temperature data and
also in adjustments for urbanisation over
the land. Over both the last 140 years
and 100 years, the best estimate is that
the global average surface temperature
has increased by 0.6 0.2°C.
(b) Additionally, the year by year (blue
curve) and 50 year average (black
curve) variations of the average surface
temperature of the Northern Hemisphere
for the past 1000 years have been
reconstructed from “proxy” data
calibrated against thermometer data (see
list of the main proxy data in the
diagram). The 95% confidence range in
the annual data is represented by the
grey region. These uncertainties increase
in more distant times and are always
much larger than in the instrumental
record due to the use of relatively sparse
proxy data. Nevertheless the rate and
duration of warming of the 20th century
has been much greater than in any of
the previous nine centuries. Similarly, it
is likely7
that the 1990s have been the
warmest decade and 1998 the warmest
year of the millennium.
[Based upon (a) Chapter 2, Figure 2.7c
and (b) Chapter 2, Figure 2.20]
3
1860 1880 1900 1920 1940 1960 1980 2000
Year
Departuresintemperature(°
fromthe1961to1990avera
Departuresintemperature(°C)
fromthe1961to1990average
(b) the past 1,000 years
NORTHERN HEMISPHERE
Data from thermometers (red) and from tree rings,
corals, ice cores and historical records (blue).
1000 1200 1400 1600 1800 2000
Year
−1.0
−0.5
0.0
0.5
Data from thermometers.
−0.8
−0.4
0.0
0.4](https://image.slidesharecdn.com/winningwaysforyourvisualizationplaysmarkgrundland-140328093347-phpapp01/85/Winning-Ways-for-your-Visualization-Plays-by-Mark-Grundland-9-320.jpg)
![How can we display our data
without distorting the truth?
Hockey stick graph was arguably the most controversial
chart in science: our future depends on how we read it.
10!
w
http://www.ipcc.ch/pdf/climate-changes-2001/synthesis-syr/english/summary-policymakers.pdf!
line, a filtered annual curve suppressing
fluctuations below near decadal
time-scales). There are uncertainties in
the annual data (thin black whisker
bars represent the 95% confidence
range) due to data gaps, random
instrumental errors and uncertainties,
uncertainties in bias corrections in the
ocean surface temperature data and
also in adjustments for urbanisation over
the land. Over both the last 140 years
and 100 years, the best estimate is that
the global average surface temperature
has increased by 0.6 0.2°C.
(b) Additionally, the year by year (blue
curve) and 50 year average (black
curve) variations of the average surface
temperature of the Northern Hemisphere
for the past 1000 years have been
reconstructed from “proxy” data
calibrated against thermometer data (see
list of the main proxy data in the
diagram). The 95% confidence range in
the annual data is represented by the
grey region. These uncertainties increase
in more distant times and are always
much larger than in the instrumental
record due to the use of relatively sparse
proxy data. Nevertheless the rate and
duration of warming of the 20th century
has been much greater than in any of
the previous nine centuries. Similarly, it
is likely7
that the 1990s have been the
warmest decade and 1998 the warmest
year of the millennium.
[Based upon (a) Chapter 2, Figure 2.7c
and (b) Chapter 2, Figure 2.20]
3
1860 1880 1900 1920 1940 1960 1980 2000
Year
Departuresintemperature(°
fromthe1961to1990avera
Departuresintemperature(°C)
fromthe1961to1990average
(b) the past 1,000 years
NORTHERN HEMISPHERE
Data from thermometers (red) and from tree rings,
corals, ice cores and historical records (blue).
1000 1200 1400 1600 1800 2000
Year
−1.0
−0.5
0.0
0.5
Data from thermometers.
−0.8
−0.4
0.0
0.4
Figure 1: Variations of the Earth’s
surface temperature over the last
140 years and the last millennium.
(a) The Earth’s surface temperature is
shown year by year (red bars) and
approximately decade by decade (black
line, a filtered annual curve suppressing
fluctuations below near decadal
time-scales). There are uncertainties in
the annual data (thin black whisker
bars represent the 95% confidence
range) due to data gaps, random
instrumental errors and uncertainties,
uncertainties in bias corrections in the
ocean surface temperature data and
also in adjustments for urbanisation over
the land. Over both the last 140 years
and 100 years, the best estimate is that
the global average surface temperature
has increased by 0.6 0.2°C.
(b) Additionally, the year by year (blue
curve) and 50 year average (black
curve) variations of the average surface
temperature of the Northern Hemisphere
for the past 1000 years have been
reconstructed from “proxy” data
calibrated against thermometer data (see
list of the main proxy data in the
diagram). The 95% confidence range in
the annual data is represented by the
grey region. These uncertainties increase
in more distant times and are always
much larger than in the instrumental
record due to the use of relatively sparse
proxy data. Nevertheless the rate and
duration of warming of the 20th century
has been much greater than in any of
the previous nine centuries. Similarly, it
is likely7
that the 1990s have been the
warmest decade and 1998 the warmest
year of the millennium.
[Based upon (a) Chapter 2, Figure 2.7c
and (b) Chapter 2, Figure 2.20]
3
1860 1880 1900 1920 1940 1960 1980 2000
Year
Departuresintemperature(°C)
fromthe1961to1990average
Departuresintemperature(°C)
fromthe1961to1990average
Variations of the Earth's surface temperature for:
(a) the past 140 years
(b) the past 1,000 years
GLOBAL
NORTHERN HEMISPHERE
Data from thermometers (red) and from tree rings,
corals, ice cores and historical records (blue).
1000 1200 1400 1600 1800 2000
Year
−1.0
−0.5
0.0
0.5
Data from thermometers.
−0.8
−0.4
0.0
0.4
0.8
1860 1880 1900 1920 1940 1960 1980 2000
Year
Departuresinte
fromthe1961to
Departuresintemperature(°C)
fromthe1961to1990average
(b) the past 1,000 years
NORTHERN HEMISPHERE
Data from thermometers (red) and from tree rings,
corals, ice cores and historical records (blue).
1000 1200 1400 1600 1800 2000
Year
−1.0
−0.5
0.0
0.5
Data from thermometers.
−0.8
−0.4
0.0](https://image.slidesharecdn.com/winningwaysforyourvisualizationplaysmarkgrundland-140328093347-phpapp01/85/Winning-Ways-for-your-Visualization-Plays-by-Mark-Grundland-10-320.jpg)
![How can we display our data
without distorting the truth?
The visualization design decisions we make affect which
interpretations of the data are facilitated or impeded.
11!
w http://www.takepart.com/an-inconvenient-truth/film!
http://www.ipcc.ch/pdf/climate-changes-2001/synthesis-syr/english/summary-policymakers.pdf!
Figure 1: Variations of the Earth’s
surface temperature over the last
140 years and the last millennium.
(a) The Earth’s surface temperature is
shown year by year (red bars) and
approximately decade by decade (black
line, a filtered annual curve suppressing
fluctuations below near decadal
time-scales). There are uncertainties in
the annual data (thin black whisker
bars represent the 95% confidence
range) due to data gaps, random
instrumental errors and uncertainties,
uncertainties in bias corrections in the
ocean surface temperature data and
also in adjustments for urbanisation over
the land. Over both the last 140 years
and 100 years, the best estimate is that
the global average surface temperature
has increased by 0.6 0.2°C.
(b) Additionally, the year by year (blue
curve) and 50 year average (black
curve) variations of the average surface
temperature of the Northern Hemisphere
for the past 1000 years have been
reconstructed from “proxy” data
calibrated against thermometer data (see
list of the main proxy data in the
diagram). The 95% confidence range in
the annual data is represented by the
grey region. These uncertainties increase
in more distant times and are always
much larger than in the instrumental
record due to the use of relatively sparse
proxy data. Nevertheless the rate and
duration of warming of the 20th century
has been much greater than in any of
the previous nine centuries. Similarly, it
is likely7
that the 1990s have been the
warmest decade and 1998 the warmest
year of the millennium.
[Based upon (a) Chapter 2, Figure 2.7c
and (b) Chapter 2, Figure 2.20]
3
1860 1880 1900 1920 1940 1960 1980 2000
Year
Departuresintemperature(°C)
fromthe1961to1990average
Departuresintemperature(°C)
fromthe1961to1990average
Variations of the Earth's surface temperature for:
(a) the past 140 years
(b) the past 1,000 years
GLOBAL
NORTHERN HEMISPHERE
Data from thermometers (red) and from tree rings,
corals, ice cores and historical records (blue).
1000 1200 1400 1600 1800 2000
Year
−1.0
−0.5
0.0
0.5
Data from thermometers.
−0.8
−0.4
0.0
0.4
0.8
1860 1880 1900 1920 1940 1960 1980 2000
Year
Departuresinte
fromthe1961to
Departuresintemperature(°C)
fromthe1961to1990average
(b) the past 1,000 years
NORTHERN HEMISPHERE
Data from thermometers (red) and from tree rings,
corals, ice cores and historical records (blue).
1000 1200 1400 1600 1800 2000
Year
−1.0
−0.5
0.0
0.5
Data from thermometers.
−0.8
−0.4
0.0](https://image.slidesharecdn.com/winningwaysforyourvisualizationplaysmarkgrundland-140328093347-phpapp01/85/Winning-Ways-for-your-Visualization-Plays-by-Mark-Grundland-11-320.jpg)
Winning Ways for your Visualization Plays by Mark Grundland
- 4. Visualization is old as art but it is just getting started “I am here” Hand cloud invented 35,000 B.C. 4! w http://en.wikipedia.org/wiki/Cueva_de_las_Manos! http://kristybuddeneducation.com/2012/07/ and http://www.tagxedo.com/app.html!
- 5. Visualization is old as art but it is just getting started “I am here” Hand cloud invented 35,000 B.C. 5! w http://en.wikipedia.org/wiki/Cueva_de_las_Manos! http://kristybuddeneducation.com/2012/07/ and http://www.tagxedo.com/app.html! “I blog here” Word cloud invented 1992 A.D.
- 6. Seeing the pattern in the data can change how we view our world Modern epidemiology started with plotting dots on a map. 6! w http://en.wikipedia.org/wiki/1854_Broad_Street_cholera_outbreak!
- 7. Modern epidemiology started with plotting dots on a map. 7! w http://en.wikipedia.org/wiki/1854_Broad_Street_cholera_outbreak! In 1854, a cholera outbreak in Soho killed over 600 people. John Snow plotted the locations of the deaths to show that they were clustered around the neighborhood water pump. Seeing the pattern in the data can change how we view our world
- 8. Communicating data effectively can change what we do with our world Hockey stick graph was arguably the most controversial chart in science: our future depends on how we read it. 8! w http://www.ipcc.ch/pdf/climate-changes-2001/synthesis-syr/english/summary-policymakers.pdf! line, a filtered annual curve suppressing fluctuations below near decadal time-scales). There are uncertainties in the annual data (thin black whisker bars represent the 95% confidence range) due to data gaps, random instrumental errors and uncertainties, uncertainties in bias corrections in the ocean surface temperature data and also in adjustments for urbanisation over the land. Over both the last 140 years and 100 years, the best estimate is that the global average surface temperature has increased by 0.6 0.2°C. (b) Additionally, the year by year (blue curve) and 50 year average (black curve) variations of the average surface temperature of the Northern Hemisphere for the past 1000 years have been reconstructed from “proxy” data calibrated against thermometer data (see list of the main proxy data in the diagram). The 95% confidence range in the annual data is represented by the grey region. These uncertainties increase in more distant times and are always much larger than in the instrumental record due to the use of relatively sparse proxy data. Nevertheless the rate and duration of warming of the 20th century has been much greater than in any of the previous nine centuries. Similarly, it is likely7 that the 1990s have been the warmest decade and 1998 the warmest year of the millennium. [Based upon (a) Chapter 2, Figure 2.7c and (b) Chapter 2, Figure 2.20] 3 1860 1880 1900 1920 1940 1960 1980 2000 Year Departuresintemperature(° fromthe1961to1990avera Departuresintemperature(°C) fromthe1961to1990average (b) the past 1,000 years NORTHERN HEMISPHERE Data from thermometers (red) and from tree rings, corals, ice cores and historical records (blue). 1000 1200 1400 1600 1800 2000 Year −1.0 −0.5 0.0 0.5 Data from thermometers. −0.8 −0.4 0.0 0.4
- 9. Communicating data effectively can change what we do with our world Hockey stick graph was arguably the most controversial chart in science: our future depends on how we read it. 9! w http://www.ipcc.ch/pdf/climate-changes-2001/synthesis-syr/english/summary-policymakers.pdf! http://www.amazon.com/dp/023115254X/ and http://www.amazon.com/dp/1906768358/! line, a filtered annual curve suppressing fluctuations below near decadal time-scales). There are uncertainties in the annual data (thin black whisker bars represent the 95% confidence range) due to data gaps, random instrumental errors and uncertainties, uncertainties in bias corrections in the ocean surface temperature data and also in adjustments for urbanisation over the land. Over both the last 140 years and 100 years, the best estimate is that the global average surface temperature has increased by 0.6 0.2°C. (b) Additionally, the year by year (blue curve) and 50 year average (black curve) variations of the average surface temperature of the Northern Hemisphere for the past 1000 years have been reconstructed from “proxy” data calibrated against thermometer data (see list of the main proxy data in the diagram). The 95% confidence range in the annual data is represented by the grey region. These uncertainties increase in more distant times and are always much larger than in the instrumental record due to the use of relatively sparse proxy data. Nevertheless the rate and duration of warming of the 20th century has been much greater than in any of the previous nine centuries. Similarly, it is likely7 that the 1990s have been the warmest decade and 1998 the warmest year of the millennium. [Based upon (a) Chapter 2, Figure 2.7c and (b) Chapter 2, Figure 2.20] 3 1860 1880 1900 1920 1940 1960 1980 2000 Year Departuresintemperature(° fromthe1961to1990avera Departuresintemperature(°C) fromthe1961to1990average (b) the past 1,000 years NORTHERN HEMISPHERE Data from thermometers (red) and from tree rings, corals, ice cores and historical records (blue). 1000 1200 1400 1600 1800 2000 Year −1.0 −0.5 0.0 0.5 Data from thermometers. −0.8 −0.4 0.0 0.4
- 10. How can we display our data without distorting the truth? Hockey stick graph was arguably the most controversial chart in science: our future depends on how we read it. 10! w http://www.ipcc.ch/pdf/climate-changes-2001/synthesis-syr/english/summary-policymakers.pdf! line, a filtered annual curve suppressing fluctuations below near decadal time-scales). There are uncertainties in the annual data (thin black whisker bars represent the 95% confidence range) due to data gaps, random instrumental errors and uncertainties, uncertainties in bias corrections in the ocean surface temperature data and also in adjustments for urbanisation over the land. Over both the last 140 years and 100 years, the best estimate is that the global average surface temperature has increased by 0.6 0.2°C. (b) Additionally, the year by year (blue curve) and 50 year average (black curve) variations of the average surface temperature of the Northern Hemisphere for the past 1000 years have been reconstructed from “proxy” data calibrated against thermometer data (see list of the main proxy data in the diagram). The 95% confidence range in the annual data is represented by the grey region. These uncertainties increase in more distant times and are always much larger than in the instrumental record due to the use of relatively sparse proxy data. Nevertheless the rate and duration of warming of the 20th century has been much greater than in any of the previous nine centuries. Similarly, it is likely7 that the 1990s have been the warmest decade and 1998 the warmest year of the millennium. [Based upon (a) Chapter 2, Figure 2.7c and (b) Chapter 2, Figure 2.20] 3 1860 1880 1900 1920 1940 1960 1980 2000 Year Departuresintemperature(° fromthe1961to1990avera Departuresintemperature(°C) fromthe1961to1990average (b) the past 1,000 years NORTHERN HEMISPHERE Data from thermometers (red) and from tree rings, corals, ice cores and historical records (blue). 1000 1200 1400 1600 1800 2000 Year −1.0 −0.5 0.0 0.5 Data from thermometers. −0.8 −0.4 0.0 0.4 Figure 1: Variations of the Earth’s surface temperature over the last 140 years and the last millennium. (a) The Earth’s surface temperature is shown year by year (red bars) and approximately decade by decade (black line, a filtered annual curve suppressing fluctuations below near decadal time-scales). There are uncertainties in the annual data (thin black whisker bars represent the 95% confidence range) due to data gaps, random instrumental errors and uncertainties, uncertainties in bias corrections in the ocean surface temperature data and also in adjustments for urbanisation over the land. Over both the last 140 years and 100 years, the best estimate is that the global average surface temperature has increased by 0.6 0.2°C. (b) Additionally, the year by year (blue curve) and 50 year average (black curve) variations of the average surface temperature of the Northern Hemisphere for the past 1000 years have been reconstructed from “proxy” data calibrated against thermometer data (see list of the main proxy data in the diagram). The 95% confidence range in the annual data is represented by the grey region. These uncertainties increase in more distant times and are always much larger than in the instrumental record due to the use of relatively sparse proxy data. Nevertheless the rate and duration of warming of the 20th century has been much greater than in any of the previous nine centuries. Similarly, it is likely7 that the 1990s have been the warmest decade and 1998 the warmest year of the millennium. [Based upon (a) Chapter 2, Figure 2.7c and (b) Chapter 2, Figure 2.20] 3 1860 1880 1900 1920 1940 1960 1980 2000 Year Departuresintemperature(°C) fromthe1961to1990average Departuresintemperature(°C) fromthe1961to1990average Variations of the Earth's surface temperature for: (a) the past 140 years (b) the past 1,000 years GLOBAL NORTHERN HEMISPHERE Data from thermometers (red) and from tree rings, corals, ice cores and historical records (blue). 1000 1200 1400 1600 1800 2000 Year −1.0 −0.5 0.0 0.5 Data from thermometers. −0.8 −0.4 0.0 0.4 0.8 1860 1880 1900 1920 1940 1960 1980 2000 Year Departuresinte fromthe1961to Departuresintemperature(°C) fromthe1961to1990average (b) the past 1,000 years NORTHERN HEMISPHERE Data from thermometers (red) and from tree rings, corals, ice cores and historical records (blue). 1000 1200 1400 1600 1800 2000 Year −1.0 −0.5 0.0 0.5 Data from thermometers. −0.8 −0.4 0.0
- 11. How can we display our data without distorting the truth? The visualization design decisions we make affect which interpretations of the data are facilitated or impeded. 11! w http://www.takepart.com/an-inconvenient-truth/film! http://www.ipcc.ch/pdf/climate-changes-2001/synthesis-syr/english/summary-policymakers.pdf! Figure 1: Variations of the Earth’s surface temperature over the last 140 years and the last millennium. (a) The Earth’s surface temperature is shown year by year (red bars) and approximately decade by decade (black line, a filtered annual curve suppressing fluctuations below near decadal time-scales). There are uncertainties in the annual data (thin black whisker bars represent the 95% confidence range) due to data gaps, random instrumental errors and uncertainties, uncertainties in bias corrections in the ocean surface temperature data and also in adjustments for urbanisation over the land. Over both the last 140 years and 100 years, the best estimate is that the global average surface temperature has increased by 0.6 0.2°C. (b) Additionally, the year by year (blue curve) and 50 year average (black curve) variations of the average surface temperature of the Northern Hemisphere for the past 1000 years have been reconstructed from “proxy” data calibrated against thermometer data (see list of the main proxy data in the diagram). The 95% confidence range in the annual data is represented by the grey region. These uncertainties increase in more distant times and are always much larger than in the instrumental record due to the use of relatively sparse proxy data. Nevertheless the rate and duration of warming of the 20th century has been much greater than in any of the previous nine centuries. Similarly, it is likely7 that the 1990s have been the warmest decade and 1998 the warmest year of the millennium. [Based upon (a) Chapter 2, Figure 2.7c and (b) Chapter 2, Figure 2.20] 3 1860 1880 1900 1920 1940 1960 1980 2000 Year Departuresintemperature(°C) fromthe1961to1990average Departuresintemperature(°C) fromthe1961to1990average Variations of the Earth's surface temperature for: (a) the past 140 years (b) the past 1,000 years GLOBAL NORTHERN HEMISPHERE Data from thermometers (red) and from tree rings, corals, ice cores and historical records (blue). 1000 1200 1400 1600 1800 2000 Year −1.0 −0.5 0.0 0.5 Data from thermometers. −0.8 −0.4 0.0 0.4 0.8 1860 1880 1900 1920 1940 1960 1980 2000 Year Departuresinte fromthe1961to Departuresintemperature(°C) fromthe1961to1990average (b) the past 1,000 years NORTHERN HEMISPHERE Data from thermometers (red) and from tree rings, corals, ice cores and historical records (blue). 1000 1200 1400 1600 1800 2000 Year −1.0 −0.5 0.0 0.5 Data from thermometers. −0.8 −0.4 0.0
- 12. How can we select the aspect ratio? !! Use 1:1 … Why? It is fair and square. !! Use 3:2 … Why? It is wider than taller, like a landscape photo. !! Use the golden ratio… Why? It is a most pleasing proportion found in nature and art. !! Make the average slope of all line segments 45º… Why? It is perceptually optimal for orientation discrimination. !! Minimize arc length, keeping area under the plot constant… Why? It is short, sweet, and mathematically optimal. !! Take the screen size or the widow size as given… Why? It fits, so obviously this must be what the user wants. !! Depends on the situation… Why? It depends on the story the user is meant to believe. 12! w http://vis.berkeley.edu/papers/banking/2006-Banking-InfoVis.pdf and http://vis.stanford.edu/papers/arclength-banking! http://www.manorpress.co.uk/printing%20info.html!
- 13. How can we select the map projection? Representing the Earth on a flat map must in some way distort distances, directions, angles, shapes, and/or areas. 13! w http://maps.google.com/" http://earthobservatory.nasa.gov/Features/10thAnniversary/top10.php! Mercator Projection Preserves angles but not areas
- 14. How can we select the map projection? Representing the Earth on a flat map must in some way distort distances, directions, angles, shapes, and/or areas. 14! w http://maps.google.com/" http://cdn.static-economist.com/sites/default/files/true-size-of-africa.jpg! Mercator Projection Preserves angles but not areas
- 15. How can we select the map projection? Representing the Earth on a flat map must in some way distort distances, directions, angles, shapes, and/or areas. !! Tissot indicatrix measures geometric distortion by showing how circles on the globe appear as ellipses on the map. 15! w http://wordpress.mrreid.org/2011/06/07/tissots-indicatrix/" http://www.progonos.com/furuti/MapProj/Normal/TOC/cartTOC.html! Mercator Projection Preserves angles but not areas Robinson Projection Nearly preserves areas but not angles
- 16. How can we select the map projection? Representing the Earth on a flat map must in some way distort distances, directions, angles, shapes, and/or areas. !! Cartograms distort the size and shape of regions in order to make their area proportional to a given variable of interest. !! Computed using density diffusion or cellular automata. 16! w http://www.worldmapper.org/! http://arxiv.org/pdf/physics/0401102v1.pdf! Land Mass Equal area cartogram Land Mass Equal area cartogram
- 17. How can we select the map projection? Representing the Earth on a flat map must in some way distort distances, directions, angles, shapes, and/or areas. !! Cartograms distort the size and shape of regions in order to make their area proportional to a given variable of interest !! Computed using density diffusion or cellular automata 17! w http://www.worldmapper.org/! http://arxiv.org/pdf/physics/0401102v1.pdf! Population Equal area cartogram GDP Wealth Equal area cartogram
- 18. How can we visually represent values? Numeric values express absolute magnitudes but visual perception makes relative judgments. !! Position !! Shape !! Length !! Orientation !! Area and volume !! Hue, saturation, brightness !! Texture and transparency !! Alignment and proximity !! Containment and connection !! Labels and glyphs !! Motion and flicker 18! w ! http://www2.parc.com/istl/projects/uir/publications/items/UIR-1986-02-Mackinlay-TOG-Automati! http://visualfunhouse.com/altered_reality/14-perfect-shot-photography-illusions.html!
- 19. How can we visually represent values? Numeric values express absolute magnitudes but visual perception makes relative judgments, not very well. Which is bigger? And by how much? 19! w http://michaelbach.de/ot/!
- 20. How can we visually represent values? Numeric values express absolute magnitudes but visual perception makes relative judgments, not very well. Which is bigger? And by how much? 20! w http://michaelbach.de/ot/! +20%
- 21. How can we visually represent values? Numeric values express absolute magnitudes but visual perception makes relative judgments, not very well. Equity Market Heat Map 21! w http://www.marketwatch.com/tools/stockresearch/marketmap!
- 22. How can we visually represent values? Numeric values express absolute magnitudes but visual perception makes relative judgments, not very well. Which is bigger? And by how much? 22! w http://en.wikipedia.org/wiki/Ebbinghaus_illusion! And by how much?
- 23. How can we visually represent values? Numeric values express absolute magnitudes but visual perception makes relative judgments, not very well. Which is bigger? And by how much? 23! w http://en.wikipedia.org/wiki/Ebbinghaus_illusion! And by how much?
- 24. How can we visually represent values? Numeric values express absolute magnitudes but visual perception makes relative judgments, not very well. Which is bigger? And by how much? 24! w http://en.wikipedia.org/wiki/Ebbinghaus_illusion! And by how much?
- 25. How can we visually represent values? Numeric values express absolute magnitudes but visual perception makes relative judgments, not very well. UK Government Spending Bubble Chart 25! w http://www.theguardian.com/news/datablog/2010/oct/18/government-spending-department-2009-10!
- 26. How can we visually represent values? Numeric values express absolute magnitudes but visual perception makes relative judgments, not very well. Which is brighter? And by how much? 26! w http://www.handprint.com/HP/WCL/tech13.html!
- 27. How can we visually represent values? Numeric values express absolute magnitudes but visual perception makes relative judgments, not very well. Which is brighter? And by how much? 27! w http://www.handprint.com/HP/WCL/tech13.html! +36%
- 28. How can we visually represent values? Numeric values express absolute magnitudes but visual perception makes relative judgments, not very well. US Presidential Election Map 28! w http://bsumaps.blogspot.co.uk/2012/11/cartographic-election-resources.html!
- 29. How can we visually represent values? Numeric values express absolute magnitudes but visual perception makes relative judgments, not very well. US Presidential Election Map 29! w http://statchatva.org/2013/01/14/lower-turnout-in-2012-makes-the-case-for-political-realignment-in-2008/!
- 30. How can we visually represent values? Help users by labeling data and adding trend indicators. US Presidential Election Map 30! w http://www.theguardian.com/news/datablog/2012/nov/07/us-elections-2012-results-map!
- 31. How can we visually represent values? Help users by labeling data and adding trend indicators. Online Ad Campaign Performance Bar Chart 31! w http://www.FunctionalElegance.com!
- 32. How can we visually represent values? Help users by labeling data and adding trend indicators. Online Ad Campaign Performance Traffic Light Bar Chart 32! w http://www.FunctionalElegance.com!
- 33. !! Hue order is not obvious. !! Hue changes make edges. !! Yellows make highlights. !! Detail is harder to see. !! Eyes are more sensitive to brightness than hues. How can we visually represent values? Perceptually uniform color gradients for continuous values. Avoid rainbow color maps: 33! w http://blog.visual.ly/rainbow-color-scales/! http://researchweb.watson.ibm.com/people/l/lloydt/color/color.HTM!
- 34. How can we visually represent values? Use transparency to overlay information layers. !! Normally transparent layers are composited using linear interpolation, an averaging operation that reduces variation. !! Blending by linear interpolation can result in reduced contrast, dull colors, detail loss, and a lack of selective emphasis. Satellite Map Overlay 34! w http://lsntap.org/blogs/gis-mapping-part-i!
- 35. How can we visually represent values? Use transparency to overlay information layers. !! Apply image blending operators that are designed to produce composite images that preserve key visual characteristics of their components: contrast, color, detail, and salience. 35! w http://www.FunctionalElegance.com/Portfolio/Publications.html!
- 36. How can we visually represent values? Use transparency to overlay information layers. !! Apply image blending operators that are designed to produce composite images that preserve key visual characteristics of their components: contrast, color, detail, and salience. 36! w http://www.FunctionalElegance.com/Portfolio/Publications.html!
- 37. How can we visually represent values? Use transparency to overlay information layers. !! Render the arc of each node connection in order of decreasing length using a color gradient that emphasizes to short links. Facebook Friend Connection Map 37! w http://paulbutler.org/archives/visualizing-facebook-friends/!
- 38. How can we visually represent values? Familiar visual metaphors make interpretation easier. SnapShot News Analysis Tool 38! w http://www.grapeshot.co.uk/snapshot/snapshot.html! http://www.FunctionalElegance.com!
- 39. How can we visually represent values? Familiar visual metaphors make interpretation easier. SnapShot News Radar 39! w http://www.grapeshot.co.uk/snapshot/snapshot.html! http://www.FunctionalElegance.com!
- 40. How can we visually represent values? Familiar visual metaphors make interpretation easier. SnapShot News Radar 40! w http://www.grapeshot.co.uk/snapshot/snapshot.html! http://www.FunctionalElegance.com!
- 41. How can we visually represent values? Familiar visual metaphors make interpretation easier. SnapShot News Radar 41! w http://www.grapeshot.co.uk/snapshot/snapshot.html! http://www.FunctionalElegance.com!
- 42. How can we visually represent connections? Use proximity to find connections in a cloud of points. !! Take a sphere of influence around each point, with radius equal to its nearest neighbor distance, and connect every pair of points whose spheres of influence intersect. Sphere of Influence Graphs Work in Rn for any Lp Metric 42! w http://dx.doi.org/10.1016/S0166-218X(02)00246-9!
- 43. How can we visually represent connections? Use tree drawing to see patterns in hierarchical data. !! Coloring each node according to its data type reveals the structure of expression trees, such as XML, JSON, and HTML. Website Home Pages as HTML Trees 43! w http://www.generatorx.no/20060528/websites-as-graphs/! http://www.flickr.com/photos/davezilla/155045024/!
- 44. How can we visually represent connections? Use word clouds that place their terms meaningfully. !! TextArc writes the sentences of a text along a circular arc and places each term according to its average position in the text. “Alice in Wonderland” TextArc 44! w http://www.textarc.org/! http://www.nytimes.com/2002/04/15/arts/15ARTS.html!
- 45. How can we visually represent connections? Use word clouds that place their terms meaningfully. !! Self-organizing maps (SOM) are neural networks, which can create knowledge maps that cluster closely associated topics. Self-organizing Map of a Mailing List 45! w http://dx.doi.org/10.1371/journal.pone.0058779! http://wiki.sugarlabs.org/go/Sugar_Labs/SOM!
- 46. How can we visually represent connections? Use graph drawing tools to see how data is connected. !! Graph drawing algorithms, such as simulated annealing or spring systems, make it easier to follow how data is related. Graph Drawing of Medical Knowledge 46! w http://cs.brown.edu/~rt/gdhandbook/! http://visualization.geblogs.com/visualization/network/!
- 47. How can we visually represent connections? Use graph drawing tools to see how data is connected. !! Hierarchical edge bundles group connections belonging to related nodes, which can be placed radially along a circle. Graph Drawing of Medical Knowledge 47! w http://www.win.tue.nl/~dholten/papers/bundles_infovis.pdf! http://visualization.geblogs.com/visualization/network/!
- 48. What is the secret to great information visualization design? Start with the user, not the data and not the graphic. !! What will this allow you to do that you can't do now? !! What difference can you observe in your business? !! What value do you expect that this will add to your business? !! What will this let your customers do that they can't do now? !! What difference can your customers observe in their business? !! What value can your customers expect that this will add? !! How does this fit in with your other strategic plans? !! How will you know that this is has been a success? !! How will you build on this success? !! So what? 48! w http://www.FunctionalElegance.com!
- 49. What does the future hold for information visualization? In an information economy, there is no shortage of information; only understanding is in short supply. !! Will interactive charts become more common as letting people play with the data drives engagement? !! Will subtly animated infographics become more common as graphic designers compete for attention? !! Will augmented reality glasses superimpose an information visualization layer on our everyday lives? !! Will cheap displays make ambient visualization ubiquitous? !! Will virtual reality and gesture interfaces have an impact? Let me know! Mark@FunctionalElegance.com 49! w http://www.FunctionalElegance.com!
- 50. Online information visualization resources Thank you for making this presentation possible! Visualization Galleries:! •! Tree Visualizations (Hans-Jörg Schulz): http://vcg.informatik.uni-rostock.de/~hs162/treeposter/poster.html! •! Time Series Visualizations (Christian Tominski & Wolfgang Aigner): http://survey.timeviz.net/! •! Visual Complexity (Manuel Lima): http://www.visualcomplexity.com/vc/! •! D3 JavaScript Visualization Library: https://github.com/mbostock/d3/wiki/Gallery! •! WebdesignerDepot.com Examples (Cameron Chapman): http://bit.ly/1nYR89L Visualization Courses: •! University of Utah (Miriah Meyer): http://www.sci.utah.edu/~miriah/cs6964/! •! University of British Columbia (Tamara Munzner): http://www.cs.ubc.ca/~tmm/courses/533-09/! •! University of California Berkeley (Michael Porath): http://blogs.ischool.berkeley.edu/i247s13/! •! University of Washington (Jeffrey Heer): https://courses.cs.washington.edu/courses/cse512/14wi/! •! Georgia Institute of Technology (John Stasko): http://www.cc.gatech.edu/~stasko/7450/syllabus.html! Visualization Tutorials: •! Storytelling with Data (Jonathan Corum): http://style.org/tapestry/! •! Visualization Analysis and Design (Tamara Munzner): http://www.cs.ubc.ca/~tmm/courses/533-11/book/! •! Principles of Information Visualization (Jessie Kennedy): http://mkweb.bcgsc.ca/vizbi/2012/! •! Information Visualization for Knowledge Discovery (Ben Shneiderman): http://bit.ly/1cw3oa2! •! Data Visualization Best Practices (Jen Underwood): http://www.slideshare.net/idigdata/! •! Data Visualization (Jan Willem Tulp): http://www.slideshare.net/janwillemtulp/! •! Information Visualization Primer (Xavier Tricoche): https://www.cs.purdue.edu/homes/cs530/new/slides/Infovis.pdf! •! Visual Techniques for Exploring Databases (Daniel A. Keim): http://www.dbs.informatik.uni-muenchen.de/~daniel/KDD97.pdf! Visualization Sites: •! Perceptual Edge (Stephen Few): http://www.perceptualedge.com/! •! Envisioning Information (Edward Tufte): http://www.edwardtufte.com/tufte/! •! Information is Beautiful (David McCandless): http://www.informationisbeautiful.net/! •! Information Aesthetics (Andrew Vande Moere): http://infosthetics.com/! •! Flowing Data (Nathan Yau): http://flowingdata.com/learning/ 50!