1
00:00:00,006 --> 00:00:05,400
Last chapter, interaction.

2
00:00:05,400 --> 00:00:12,670
Let's just start by taking a look at
a very classic data of this example.

3
00:00:12,670 --> 00:00:18,480
And this is Marey's Paris to Lyon train
schedule, and this comes from 1885.

4
00:00:18,480 --> 00:00:22,370
It's one of those sort of
masterpieces of visualization.

5
00:00:22,370 --> 00:00:26,800
And what I'd like to do is show you
how something even as marvelous as

6
00:00:26,800 --> 00:00:31,020
this can be improved with
something called interaction.

7
00:00:31,020 --> 00:00:37,090
So, in this particular diagram,
the x axis represents time and the y axis

8
00:00:37,090 --> 00:00:42,230
represents the nominal location for
top Paris, bottom Lyon.

9
00:00:42,230 --> 00:00:46,810
And what you can see is,
each train is represented by a different,

10
00:00:46,810 --> 00:00:48,760
a line of a different slope.

11
00:00:48,760 --> 00:00:54,620
So lines that are sloping to the right
are trains that go from Paris to Lyon,

12
00:00:54,620 --> 00:00:59,030
and trains that are sloping up into the
right are trains going from Lyon to Paris.

13
00:01:00,060 --> 00:01:03,710
So something as wonderful as this can be

14
00:01:03,710 --> 00:01:07,380
improved with some very
simple interaction.

15
00:01:08,570 --> 00:01:14,020
So what we've done here is this
is a code written by Mike Bostock

16
00:01:14,020 --> 00:01:19,920
in a very excellent online
graphing library called D3,

17
00:01:19,920 --> 00:01:23,650
and he's added some interactive
widgets across the top.

18
00:01:23,650 --> 00:01:29,300
And this allows us to do things like, well
let's just turn off the southbound trains.

19
00:01:30,650 --> 00:01:35,450
And I'm only interested in not
the everyday but just Saturdays.

20
00:01:36,630 --> 00:01:39,650
And I'm only interested
in the bullet trains.

21
00:01:39,650 --> 00:01:41,380
Oh look,
there's actually none on Saturday.

22
00:01:42,820 --> 00:01:46,420
So I wonder,
I'll have to look at weekdays.

23
00:01:46,420 --> 00:01:48,100
Looks like there's plenty during weekdays.

24
00:01:49,180 --> 00:01:56,450
So you're ability to actually change
what data are being represented

25
00:01:56,450 --> 00:02:02,100
from the entirety of the data
set gives us an incredible power

26
00:02:02,100 --> 00:02:07,760
to see patterns that are ones
that aren't necessarily

27
00:02:07,760 --> 00:02:13,020
visible in the encoding of
the data set in its entirety.

28
00:02:14,460 --> 00:02:20,210
Let's look at some other examples,
and see how our ability to interact

29
00:02:20,210 --> 00:02:25,750
with data also give us an incredible
power to learn things about patterns.

30
00:02:26,900 --> 00:02:28,980
So here is something
called the zip decoder.

31
00:02:28,980 --> 00:02:30,430
It's made by Ben Fry.

32
00:02:30,430 --> 00:02:34,330
It's a wonderfully simple data
visualization where what Ben has done is

33
00:02:34,330 --> 00:02:42,770
he's mapped every, unique postal code
to a dot, in latitude and longitude.

34
00:02:42,770 --> 00:02:47,370
And what you can do is, by clicking on the
map and you enter, everybody just enters

35
00:02:47,370 --> 00:02:54,050
their own zip code, and they're able to
see the, how the zip codes form a shape.

36
00:02:54,050 --> 00:02:56,490
And so for example,
I come from Long Island,

37
00:02:56,490 --> 00:03:02,350
New York, so I would type one, and you
can see here that this is the Northeast.

38
00:03:02,350 --> 00:03:08,200
One, that's Long Island, and
then seven, four, seven,

39
00:03:08,200 --> 00:03:13,510
and you can see we've slowly,
gradually reduced the amount of noise.

40
00:03:13,510 --> 00:03:16,450
But, of course you can
enter whatever you want and

41
00:03:16,450 --> 00:03:19,350
you can see here quickly
that if you just type nine,

42
00:03:19,350 --> 00:03:24,850
that this is the West Coast, and
that eight is slightly more inward.

43
00:03:24,850 --> 00:03:27,770
Seven is Texas in the South.

44
00:03:27,770 --> 00:03:30,190
Six, five, four, so

45
00:03:30,190 --> 00:03:35,410
you can see there are different patterns,
that in the entirety aren't visible.

46
00:03:35,410 --> 00:03:37,360
But by our ability to interact,

47
00:03:37,360 --> 00:03:41,060
what we've really done here is
something called dynamic filtering.

48
00:03:41,060 --> 00:03:48,240
And what we're doing is we're maintaining
the representation, but we are removing or

49
00:03:48,240 --> 00:03:53,680
highlighting data depending on
commands that we give the computer.

50
00:03:53,680 --> 00:03:58,460
And so here when I'm saying to
show me just these, what it's

51
00:03:58,460 --> 00:04:02,570
really doing is making more salient the
data that I'm interested in looking at.

52
00:04:05,700 --> 00:04:12,770
Another example of this type of filtering;
here is the different types of travel

53
00:04:14,225 --> 00:04:19,235
you know, airline travel
between where is it exactly?

54
00:04:19,235 --> 00:04:25,100
[LAUGH]
Between somewhere and somewhere else.

55
00:04:25,100 --> 00:04:31,530
But the main point is that what you can
see here is the top is a one chart,

56
00:04:31,530 --> 00:04:36,070
a histogram, showing arrivals by
time of day and this range of date.

57
00:04:36,070 --> 00:04:42,010
And what we can do here is if we increase
the date, what you can see is we ad and

58
00:04:42,010 --> 00:04:46,790
you'll see the time of day that
histogram is getting the shape,

59
00:04:46,790 --> 00:04:50,590
the shape of the histogram is
changing based on the actual data.

60
00:04:50,590 --> 00:04:54,900
And we're changing the entirety of
the data that's being included, and

61
00:04:54,900 --> 00:05:00,770
then what you can do is using the cursor,
select just a sub sample of these below.

62
00:05:00,770 --> 00:05:06,230
So now what I'm doing is looking at
data between January and February, and

63
00:05:06,230 --> 00:05:11,250
just at, you know, between six and
10 in the morning.

64
00:05:11,250 --> 00:05:14,180
And I can now look at
different flights here.

65
00:05:14,180 --> 00:05:19,500
So if I'm interested in seeing, oh well
what about flights that are, you know,

66
00:05:19,500 --> 00:05:27,420
in February and that are in the afternoon.

67
00:05:27,420 --> 00:05:33,110
We can see here that this is
the list of flights below and

68
00:05:33,110 --> 00:05:36,560
up top,
we can see the up time the arrival delay.

69
00:05:36,560 --> 00:05:39,460
So for example if I want to pick flight,
of these flights,

70
00:05:39,460 --> 00:05:44,050
the ones that are the most on time,
I can sub select again.

71
00:05:44,050 --> 00:05:49,070
So this ability to filter, right,
the underlying representation named is

72
00:05:49,070 --> 00:05:52,960
the same, and
the visual presentation is the same.

73
00:05:52,960 --> 00:05:58,620
All we're adding is this layer that's
saying, show me some of the data,

74
00:05:58,620 --> 00:06:00,500
but not all of the data.

75
00:06:00,500 --> 00:06:07,140
And, because you're showing the data, that
maps to the questions that you're asking,

76
00:06:07,140 --> 00:06:10,660
you're actually able to understand
a great deal about the data.

77
00:06:12,260 --> 00:06:17,390
Another wonderful example of this
kind of dynamic filtering is a data

78
00:06:17,390 --> 00:06:23,610
visualization called five years of info
aesthetic which takes the info aesthetic

79
00:06:23,610 --> 00:06:30,330
visualization blog and every single
one of the features from that blog.

80
00:06:30,330 --> 00:06:35,190
And so for example, I can and you can see,
it organizes them by category.

81
00:06:35,190 --> 00:06:37,020
And if I click on one,

82
00:06:37,020 --> 00:06:42,390
it will show me what categories that
particular visualization belong to, and it

83
00:06:42,390 --> 00:06:47,440
will actually get an image that represents
it and show me the ability to go there.

84
00:06:48,460 --> 00:06:52,220
And now if I was interested in sub
sampling I could actually say well how

85
00:06:52,220 --> 00:06:56,690
about the, this particular visualization
well this one is an aesthetic or

86
00:06:56,690 --> 00:06:57,960
info graphic.

87
00:06:57,960 --> 00:07:00,030
Oh, well I see how this works.

88
00:07:00,030 --> 00:07:02,860
What if I'm only interested
in seeing the ones with art.

89
00:07:02,860 --> 00:07:08,310
And so what happens is, the art becomes
the top level representation, and

90
00:07:08,310 --> 00:07:11,940
all of the, the headlines that don't
include art become diminished.

91
00:07:13,470 --> 00:07:19,870
So, now I'm actually able to sample,
through time, the different visualization,

92
00:07:19,870 --> 00:07:25,720
the different features that include
art as their top level component.

93
00:07:26,840 --> 00:07:31,600
So what you can see here is
this is a fascinating way to,

94
00:07:31,600 --> 00:07:37,190
by using this idea of dynamic filtering
and changing what's being presented,

95
00:07:37,190 --> 00:07:42,120
you're able to give the individual
this capability to see into the data.

96
00:07:43,150 --> 00:07:47,450
So visualization excuse me,
interaction can be a hard thing to fathom.

97
00:07:47,450 --> 00:07:52,040
So I look to Ben Shneiderman's
example of what is it that we

98
00:07:52,040 --> 00:07:57,340
should do in order to make interaction
with visualization as simple as possible.

99
00:07:57,340 --> 00:08:00,650
And he has three wonderful guidelines.

100
00:08:00,650 --> 00:08:05,990
Overview first, zoom and
filter, details-on-demand.

101
00:08:05,990 --> 00:08:07,960
So, let's go back to look at this.

102
00:08:07,960 --> 00:08:13,540
When we start with the overview,
what we're able to do is see the entirety.

103
00:08:13,540 --> 00:08:19,430
Wow, there are probably 500 of these,
you can see the number and

104
00:08:19,430 --> 00:08:22,809
the distribution of,
over subjects on the right.

105
00:08:24,270 --> 00:08:26,370
Overview is what you get
when you first come.

106
00:08:27,960 --> 00:08:29,508
Zoom and filter.

107
00:08:29,508 --> 00:08:36,200
So, I'm able, I'm able to filter and

108
00:08:36,200 --> 00:08:42,120
I'm able to zoom by seeing, well actually
I'm not able to zoom in this one, but.

109
00:08:42,120 --> 00:08:47,350
And then details on demand, so
when I'm interested in finding out more,

110
00:08:47,350 --> 00:08:52,380
I can enter very specifically,
I want to see this one, and

111
00:08:52,380 --> 00:08:57,250
then it gives you some of
the much more detailed layer.

112
00:08:57,250 --> 00:09:00,290
So, you see,
by layering the visualization and

113
00:09:00,290 --> 00:09:05,380
allowing people to stand back and
see the big picture or to zoom in, and

114
00:09:05,380 --> 00:09:09,830
see the individual data points
you're taking advantage of one of

115
00:09:09,830 --> 00:09:15,815
the things that interaction can do that
other types of visualization simply can't.

116
00:09:15,815 --> 00:09:21,450
Looking so last

117
00:09:21,450 --> 00:09:25,270
something that I think is just a joy I
recommend everybody doing is kind of fun.

118
00:09:25,270 --> 00:09:29,610
It's something that was come
up by Martin Wattenberg and

119
00:09:29,610 --> 00:09:32,260
it's basically called
the baby name generator.

120
00:09:32,260 --> 00:09:37,680
Or it really is, is a statistical
evaluation of names mentioned

121
00:09:37,680 --> 00:09:42,200
in the United States Census
from the 1880s to today.

122
00:09:42,200 --> 00:09:47,460
And it shows the top 2,000
most popular names and

123
00:09:47,460 --> 00:09:52,050
what you can do is track the historical
popularity of different names.

124
00:09:52,050 --> 00:09:57,130
So I'll write here for example if I
want to look at Scott, I would type in S,

125
00:09:57,130 --> 00:10:01,490
it'll filter down on different
names that start from S.

126
00:10:01,490 --> 00:10:05,560
And what you can see here is, Scott wasn't
actually very popular in the United States

127
00:10:05,560 --> 00:10:10,280
until the 1930s,
achieved peak popularity in the 50s and

128
00:10:10,280 --> 00:10:14,650
60s, and then is now actually
not a very common name at all.

129
00:10:16,660 --> 00:10:22,290
And then if you look at something like L,
you can see names like Larry,

130
00:10:22,290 --> 00:10:25,530
more classic American names or

131
00:10:25,530 --> 00:10:29,260
a name like Linda which had a burst
in popularity between the 30s and

132
00:10:29,260 --> 00:10:34,860
the 60s and is now represents a much
smaller percentage of the population.

133
00:10:36,440 --> 00:10:38,240
Fine interactive date visualizations.

134
00:10:38,240 --> 00:10:44,650
I think which we'll find
is that interaction is

135
00:10:44,650 --> 00:10:47,940
one of the hardest parts of
data visualization to master.

136
00:10:48,940 --> 00:10:54,920
But it's also one of the richest,
because it's the part that allows

137
00:10:54,920 --> 00:11:01,440
you to dive into a visualization and
extract answers, that

138
00:11:01,440 --> 00:11:06,660
are potentially harder to gather
when looking at your data

139
00:11:06,660 --> 00:11:12,390
from a single perspective is
challenging because of its density or

140
00:11:12,390 --> 00:11:14,630
perhaps you're not even sure what
you're looking for right yet.

141
00:11:16,310 --> 00:11:21,120
So, I recommend, tremendous,
well actually one other thing.

142
00:11:21,120 --> 00:11:25,472
There's I'll just put
this teaser at the end.

143
00:11:25,472 --> 00:11:30,600
Last year myself and
a number of colleagues gave a number of

144
00:11:30,600 --> 00:11:34,360
keynote lectures on the subject
of What is data visualization.

145
00:11:34,360 --> 00:11:39,920
I think you'll find that there's
a really wonderful cross section

146
00:11:39,920 --> 00:11:43,160
of thoughts about what visualization is.

147
00:11:44,300 --> 00:11:46,560
If you look at the handouts
from the lectures,

148
00:11:46,560 --> 00:11:49,550
each of the links will be in there,
so you'll be able to click them and

149
00:11:49,550 --> 00:11:52,430
explore each of the talks
is about an hour.

150
00:11:52,430 --> 00:11:53,240
So enjoy them.

151
00:11:53,240 --> 00:11:58,000
I hope this was interesting,
and, you know, think visually.

152
00:11:59,030 --> 00:11:59,670
Thanks very much