﻿1
00:00:00,133 --> 00:00:02,566
Hello and welcome to this art tutorial.

2
00:00:02,566 --> 00:00:04,633
So now it's time for the fun part.

3
00:00:04,633 --> 00:00:08,166
I can't wait to show you the training
set results and the test set results.

4
00:00:08,633 --> 00:00:11,633
So let's start to build the graphs.

5
00:00:12,266 --> 00:00:16,166
So the first thing we're going to do
is to import the ggplot2

6
00:00:16,166 --> 00:00:19,166
library, which is a very good way
of plotting something in R.

7
00:00:19,500 --> 00:00:21,466
So I'm going to import it.

8
00:00:21,466 --> 00:00:24,466
But before importing this library
there might be some of you

9
00:00:24,500 --> 00:00:26,166
who are starting R for the first time.

10
00:00:26,166 --> 00:00:29,633
So if we go to our packages here,
you can see that

11
00:00:29,700 --> 00:00:32,633
the ggplot2 library is here on my RStudio.

12
00:00:32,633 --> 00:00:34,633
But it might not be the case for you
because

13
00:00:34,633 --> 00:00:38,633
and that would mean that the ggplot2
library is not installed on your R.

14
00:00:38,966 --> 00:00:42,666
So just for those of you for which it's
the case, I'm going to install

15
00:00:44,000 --> 00:00:45,500
ggplot2 library.

16
00:00:45,500 --> 00:00:48,633
So we have to use this command
install packages here.

17
00:00:49,166 --> 00:00:53,033
And then in parenthesis we add ggplot2

18
00:00:53,600 --> 00:00:57,300
g plot two.

19
00:00:57,600 --> 00:01:00,766
Then we select this press
command and control plus enter to execute.

20
00:01:01,633 --> 00:01:04,633
And here and as you can see
it's installing the library.

21
00:01:04,833 --> 00:01:06,100
And that's done.

22
00:01:06,100 --> 00:01:10,866
So that's the message you should be
obtaining when you install the package.

23
00:01:11,466 --> 00:01:13,700
Now it should be appearing
in your library.

24
00:01:13,700 --> 00:01:17,666
So mine is here and we are ready
to start plotting the graph.

25
00:01:18,600 --> 00:01:21,100
I'm just going to put that in comment.

26
00:01:21,100 --> 00:01:23,200
So command shift plus C.

27
00:01:23,200 --> 00:01:23,866
Now it's incoming

28
00:01:23,866 --> 00:01:26,866
because once it's installed
we don't need to install it again.

29
00:01:27,666 --> 00:01:27,966
Okay.

30
00:01:27,966 --> 00:01:32,100
So now let's start to plot our training
set results graph.

31
00:01:32,700 --> 00:01:36,333
So okay so first thing we need to do
as you can see the ggplot

32
00:01:36,333 --> 00:01:38,266
two library is not selected here.

33
00:01:38,266 --> 00:01:39,933
So we will need to select it.

34
00:01:39,933 --> 00:01:43,633
But we want to automate all this
if we want to use simple linear

35
00:01:43,633 --> 00:01:45,300
regression on another data set.

36
00:01:45,300 --> 00:01:47,000
So we will just type here library.

37
00:01:48,366 --> 00:01:50,666
Then in
parentheses the name of the library.

38
00:01:50,666 --> 00:01:53,666
Not in quotes actually did ggplot2.

39
00:01:53,800 --> 00:01:55,833
And as you can see if I select this line

40
00:01:55,833 --> 00:01:59,466
and execute here,
you see that it's not selected.

41
00:01:59,466 --> 00:02:04,133
And if I execute it selected here
we just have a suggestion

42
00:02:04,133 --> 00:02:08,033
to type suppress package started messages
to eliminate package started messages.

43
00:02:08,533 --> 00:02:10,000
Well that's not very important.

44
00:02:10,000 --> 00:02:13,000
You can execute that if you don't want
the package started messages.

45
00:02:13,400 --> 00:02:14,866
But that's okay.

46
00:02:14,866 --> 00:02:17,866
So I'm just going to enlarge this

47
00:02:18,200 --> 00:02:18,566
okay.

48
00:02:18,566 --> 00:02:22,500
And now like for this course the approach
we will be taking to build

49
00:02:22,500 --> 00:02:27,366
this graph with ggplot2 is going to be
a step by step approach, because we are

50
00:02:27,366 --> 00:02:31,166
first going to plot all the observation
points in the training set.

51
00:02:31,633 --> 00:02:34,633
Then we're going to plot
the regression line.

52
00:02:34,633 --> 00:02:38,266
And then we're going to add a title
and then a label to the x axis

53
00:02:38,266 --> 00:02:39,633
and a label to the y axis.

54
00:02:39,633 --> 00:02:41,500
So you know it's going to be step by step.

55
00:02:41,500 --> 00:02:43,500
First the observation points,

56
00:02:43,500 --> 00:02:46,600
then the regression line,
then the title, then the labels.

57
00:02:46,966 --> 00:02:47,300
Okay.

58
00:02:47,300 --> 00:02:50,033
So let's start with the observation
points.

59
00:02:50,033 --> 00:02:51,233
So okay.

60
00:02:51,233 --> 00:02:54,966
So generally to introduce a plot in ggplot
we start by typing

61
00:02:54,966 --> 00:02:57,966
ggplot parenthesis.

62
00:02:58,266 --> 00:02:58,933
All right.

63
00:02:58,933 --> 00:02:59,466
And then that's

64
00:02:59,466 --> 00:03:02,966
when we start to build step by step
the little components of our graph.

65
00:03:03,266 --> 00:03:06,000
And since we're adding each component
one by one

66
00:03:06,000 --> 00:03:09,000
we are going to separate
each of these components by plus.

67
00:03:09,266 --> 00:03:13,766
So let's go for the first component
plus here I will press enter.

68
00:03:14,100 --> 00:03:16,566
And here we go for the first component.

69
00:03:16,566 --> 00:03:19,333
So the first component
are the observation points.

70
00:03:19,333 --> 00:03:22,733
So very intuitively
we are going to write geom

71
00:03:23,700 --> 00:03:26,400
underscore point.

72
00:03:26,400 --> 00:03:27,300
All right.

73
00:03:27,300 --> 00:03:30,300
So we are going to use this function
of ggplot

74
00:03:30,500 --> 00:03:34,566
to scatter plot all our observation
points of the training set.

75
00:03:35,066 --> 00:03:38,333
So the first thing that we need to do
is to precise

76
00:03:38,333 --> 00:03:41,666
which are going to be the x axis
and the y axis.

77
00:03:41,666 --> 00:03:46,766
And that's in the esthetic function
which we call x.

78
00:03:47,233 --> 00:03:51,933
So that's the function that will take as
input the x variables and the y variable.

79
00:03:51,966 --> 00:03:54,600
So the x variables is our years
experience variable.

80
00:03:54,600 --> 00:03:56,400
That means the number of years
of experience.

81
00:03:56,400 --> 00:03:58,566
That's what we want to have in the x axis.

82
00:03:58,566 --> 00:04:01,766
And the y variable will be our salaries
but the real ones

83
00:04:01,933 --> 00:04:03,866
not the predicted ones. Yet.

84
00:04:03,866 --> 00:04:07,466
So therefore we are
going to add in parentheses x

85
00:04:08,433 --> 00:04:09,366
equals.

86
00:04:09,366 --> 00:04:12,366
And then we're
going to put all our years of experience.

87
00:04:12,600 --> 00:04:13,800
But be careful.

88
00:04:13,800 --> 00:04:17,366
We want to take the years of experience
of the observations in the training set.

89
00:04:17,766 --> 00:04:22,433
So here we'll just have to specify
training set dollar

90
00:04:22,800 --> 00:04:24,700
and then years experience.

91
00:04:24,700 --> 00:04:28,500
So that means that we are taking the years
experience values

92
00:04:29,100 --> 00:04:32,100
of the observations
in the training set okay.

93
00:04:32,566 --> 00:04:35,466
And now same for Y.

94
00:04:35,466 --> 00:04:40,000
And here we do
the same only with the salaries

95
00:04:40,866 --> 00:04:43,166
which is here okay.

96
00:04:43,166 --> 00:04:46,166
And that plots all your observation
points.

97
00:04:46,233 --> 00:04:49,600
The real ones in the plot okay.

98
00:04:49,600 --> 00:04:52,600
So we could stop here,
but we would like to add a call

99
00:04:52,666 --> 00:04:56,500
because we will want to make
the distinction between the observation

100
00:04:56,500 --> 00:05:00,133
points and the regression line
to, you know, make a nice plot.

101
00:05:00,466 --> 00:05:03,533
So come here and then I'm going to press

102
00:05:03,533 --> 00:05:06,600
enter to add a new argument of geom point.

103
00:05:06,600 --> 00:05:09,133
Because this can be a little confusing
here we have

104
00:05:09,133 --> 00:05:10,966
we have a coming here and coming here.

105
00:05:10,966 --> 00:05:14,500
This coming here is just to separate the x
and the y in the esthetic function.

106
00:05:14,733 --> 00:05:16,433
And this coming here is to separate

107
00:05:16,433 --> 00:05:19,733
the first argument and the second argument
of the June point function.

108
00:05:20,100 --> 00:05:21,833
So the second argument is color.

109
00:05:23,133 --> 00:05:25,566
And let's pick red.

110
00:05:25,566 --> 00:05:28,200
I usually like to have my observation
points in red.

111
00:05:28,200 --> 00:05:30,100
But you know that's just my taste.

112
00:05:30,100 --> 00:05:33,000
If you have some other tastes,
you're welcome to, choose

113
00:05:33,000 --> 00:05:36,000
your favorite color for the observation
points.

114
00:05:36,600 --> 00:05:38,966
Okay, so that's it for the geom point.

115
00:05:38,966 --> 00:05:42,966
With this zoom point part, here
we are plotting all the observation

116
00:05:42,966 --> 00:05:45,966
points of our training set.
