1
00:00:00,025 --> 00:00:03,782
[MUSIC]

2
00:00:03,782 --> 00:00:07,583
So, now that we've took
a quick look at the dataset,

3
00:00:07,583 --> 00:00:11,290
let me show you something pretty cool.

4
00:00:11,290 --> 00:00:17,830
So, let's introduce what's
called the GraphLab Canvas.

5
00:00:19,830 --> 00:00:23,720
So this kind of type or data, graph
data takes an image data that we have.

6
00:00:23,720 --> 00:00:25,750
There's many ways you can visualize it.

7
00:00:25,750 --> 00:00:29,990
What we have inside GraphLab Create
is a data visualization companion

8
00:00:29,990 --> 00:00:33,530
that can be used to view
the data in various ways.

9
00:00:33,530 --> 00:00:38,460
So if you take any data
structure that we play with,

10
00:00:38,460 --> 00:00:40,140
any data structure GraphLab Create.

11
00:00:40,140 --> 00:00:42,433
So Take any data structure,

12
00:00:46,237 --> 00:00:53,017
In GraphLab Create and
you type .show after it.

13
00:00:53,017 --> 00:00:56,568
So if you type sf.show is going
to start the GraphLab Canvas and

14
00:00:56,568 --> 00:01:00,990
give you visualization,
first visualization of the data structure.

15
00:01:00,990 --> 00:01:02,321
In this case,
we have a very simple dataset.

16
00:01:02,321 --> 00:01:07,100
When I tab, it does show,
it gives me a link here

17
00:01:07,100 --> 00:01:11,650
to a new Web page that
shows you that dataset.

18
00:01:11,650 --> 00:01:14,260
So I'm gonna decrease
the font size a little bit.

19
00:01:14,260 --> 00:01:18,370
And you'll see that this
dataset has four columns.

20
00:01:18,370 --> 00:01:21,300
The First Name, Alex,
Malcolm, Felix, and so on.

21
00:01:21,300 --> 00:01:21,812
The Last Name.

22
00:01:21,812 --> 00:01:26,507
The Country this person lives in and
their age.

23
00:01:26,507 --> 00:01:30,984
And you see that the max age is 25,
the min age is 22,

24
00:01:30,984 --> 00:01:35,862
the median age is 23 the mean
age is 23.143 and so on.

25
00:01:35,862 --> 00:01:39,577
And if you can click on tabular view,
you can actually view the data.

26
00:01:39,577 --> 00:01:41,220
And this is pretty cool.

27
00:01:41,220 --> 00:01:44,210
If you have a million rows, it's
really hard to actually view the data.

28
00:01:44,210 --> 00:01:47,412
But this allows you to
scroll through millions,

29
00:01:47,412 --> 00:01:51,721
1 billion rows you could scroll
through with this tabular view.

30
00:01:51,721 --> 00:01:56,200
And now, if I click on any column,
for example, the age column,

31
00:01:56,200 --> 00:02:02,000
it's going to give me, sorry, I selected
something here which was a little weird.

32
00:02:02,000 --> 00:02:06,210
Let me just type reload because
it did something funny.

33
00:02:06,210 --> 00:02:08,041
A histogram, but if I type Categorical,

34
00:02:08,041 --> 00:02:10,041
it's gonna sort it by
the most common value.

35
00:02:10,041 --> 00:02:16,717
You see that the most common Value Is
a 23, there's 3 of these folks,

36
00:02:16,717 --> 00:02:21,220
and so
that's 42% of the data has age of 23.

37
00:02:21,220 --> 00:02:26,950
So we can play with Canvas and visualize
the data in various ways and get started.

38
00:02:28,000 --> 00:02:32,610
Now for the notebooks that
we're filming here, the Canvas

39
00:02:32,610 --> 00:02:37,210
popping up in a new window like this
makes it kind of awkward to go back and

40
00:02:37,210 --> 00:02:39,710
forth when we're just filming.

41
00:02:39,710 --> 00:02:44,184
So, instead of open as a new window,
what I'm going to see is how to open

42
00:02:44,184 --> 00:02:48,690
the Canvas or some of visualizations
inside of the IPhython Notebook.

43
00:02:48,690 --> 00:02:55,483
So, I'm going to call a command
called graphlab.canvas.set_target,

44
00:02:55,483 --> 00:03:02,074
and I'm gonna say that the target is
the IPython Notebook, short ipynb.

45
00:03:02,074 --> 00:03:04,370
And we're gonna do this again,
and again, and again.

46
00:03:04,370 --> 00:03:08,930
But what happens is now the visualizations
that we do are all gonna target

47
00:03:08,930 --> 00:03:10,760
this actual notebook.

48
00:03:10,760 --> 00:03:17,140
So, let's visualize, again, the age
column, but now inside our notebook.

49
00:03:17,140 --> 00:03:18,321
So here's what we're gonna do,
it's pretty cool.

50
00:03:18,321 --> 00:03:22,440
We're gonna take this s frame,
sf, and I'm gonna select.

51
00:03:22,440 --> 00:03:26,900
The square brackets allows you to select
a particular column or many columns.

52
00:03:26,900 --> 00:03:33,679
So, I'm gonna select the column age,
just one, and I'm gonna type .show on it.

53
00:03:33,679 --> 00:03:39,279
And I'm gonna tell it that
the view that I want is

54
00:03:39,279 --> 00:03:44,890
that sorted Categorical
view that we just saw.

55
00:03:44,890 --> 00:03:46,280
So, here we go.

56
00:03:46,280 --> 00:03:50,540
We just saw that 23 is
the most common age.

57
00:03:50,540 --> 00:03:53,920
People with 23 years of age,
42% of the data.

58
00:03:53,920 --> 00:04:00,210
So, now we see how to do a little bit
of visualization using GraphLab Create.

59
00:04:00,210 --> 00:04:03,047
We're gonna see several other
visualizations within GraphLab and

60
00:04:03,047 --> 00:04:04,697
within other tools during the course.

61
00:04:04,697 --> 00:04:09,379
[MUSIC]