﻿1
00:00:00,100 --> 00:00:01,000
All right, my friends.

2
00:00:01,000 --> 00:00:04,233
Are you ready for the final steps
of this implementation?

3
00:00:04,333 --> 00:00:06,966
First, we're going to visualize
the training set results.

4
00:00:06,966 --> 00:00:09,666
And second,
we will visualize the test set results.

5
00:00:09,666 --> 00:00:10,633
So let's do this.

6
00:00:10,633 --> 00:00:12,966
Let's start by creating a new code cell.

7
00:00:12,966 --> 00:00:14,066
And there we go.

8
00:00:14,066 --> 00:00:18,000
Now we just want to plot on a nice graphic
the real salaries

9
00:00:18,000 --> 00:00:20,000
compared to the predicted salaries.

10
00:00:20,000 --> 00:00:21,966
So basically we're going to have a 2D plot

11
00:00:21,966 --> 00:00:25,500
with the x axis
being the numbers of years of experience,

12
00:00:25,500 --> 00:00:30,200
you know, from 1 to 10
and the y axis being the salaries,

13
00:00:30,200 --> 00:00:33,600
you know, in the range of salaries
given in this data set.

14
00:00:34,000 --> 00:00:37,433
And so we will plot in red points
the real salaries

15
00:00:37,700 --> 00:00:41,200
and in the blue straight line,
the predicted salaries.

16
00:00:41,200 --> 00:00:41,700
Okay.

17
00:00:41,700 --> 00:00:43,166
And we will do that both

18
00:00:43,166 --> 00:00:46,933
for the predictions in the training set
and the predictions in a test set.

19
00:00:47,400 --> 00:00:48,100
Are you ready?

20
00:00:48,100 --> 00:00:49,433
Let's do this.

21
00:00:49,433 --> 00:00:52,366
We're going to first call matplotlib
as a reminder.

22
00:00:52,366 --> 00:00:54,266
It has the shortcut plt.

23
00:00:54,266 --> 00:00:55,533
Actually what we're going to call

24
00:00:55,533 --> 00:00:58,533
is exactly the pie plot module
from the matplotlib library.

25
00:00:58,533 --> 00:01:01,933
And this is
what is having the PLT shortcut.

26
00:01:02,233 --> 00:01:03,366
So there we go.

27
00:01:03,366 --> 00:01:06,933
Let's start by calling Pyplot plt.

28
00:01:07,800 --> 00:01:08,666
And then we're going to call

29
00:01:08,666 --> 00:01:12,000
a specific function from this module
which is called scatter.

30
00:01:12,700 --> 00:01:16,033
And scatter will allow us
to put the red points

31
00:01:16,033 --> 00:01:19,033
in a corresponding
to the real salaries in the 2D plot.

32
00:01:19,133 --> 00:01:23,400
Okay, so naturally what we have to input
here are the coordinates

33
00:01:23,400 --> 00:01:26,466
of these red points
representing the real salaries

34
00:01:26,733 --> 00:01:30,600
and these coordinates are nothing else
than X train and Y train.

35
00:01:30,600 --> 00:01:31,000
Right.

36
00:01:31,000 --> 00:01:35,233
Because remember the x axis will be
the numbers of years of experience.

37
00:01:35,233 --> 00:01:38,033
That's what is contained in x train.

38
00:01:38,033 --> 00:01:40,200
And the y axis will be the salaries.

39
00:01:40,200 --> 00:01:42,033
And that is what is contained in y train.

40
00:01:42,033 --> 00:01:43,566
So here and scatter here.

41
00:01:43,566 --> 00:01:45,266
We just have to input the coordinates

42
00:01:45,266 --> 00:01:48,866
first X train
for the numbers of years of experience.

43
00:01:49,133 --> 00:01:49,800
There you go.

44
00:01:49,800 --> 00:01:53,466
And then Y train for the salaries okay.

45
00:01:54,500 --> 00:01:55,233
And then we

46
00:01:55,233 --> 00:01:58,233
can add another argument which is color

47
00:01:58,400 --> 00:02:02,133
and which will allow us
to choose color of these points.

48
00:02:02,133 --> 00:02:05,066
And as we said, we're going to choose red.

49
00:02:05,066 --> 00:02:06,000
Perfect.

50
00:02:06,000 --> 00:02:09,266
Then next step
is to plot the regression line.

51
00:02:09,266 --> 00:02:13,700
So remember from the intuition lectures,
the regression line is the line

52
00:02:13,700 --> 00:02:18,433
of the predictions coming as close
as possible to the real results,

53
00:02:18,433 --> 00:02:21,833
you know, to the real salaries
and therefore the predictions.

54
00:02:21,833 --> 00:02:25,433
You know, the points corresponding
to the predicted salaries will follow

55
00:02:25,433 --> 00:02:28,600
a straight line, right,
as in a linear function.

56
00:02:29,000 --> 00:02:32,000
And therefore here
we're not going to use the scatter method.

57
00:02:32,000 --> 00:02:34,433
We're going to use the plot method

58
00:02:34,433 --> 00:02:37,233
because that's what we use
to plot the curve of a function.

59
00:02:37,233 --> 00:02:40,900
And since our function is linear,
it will actually be a straight line.

60
00:02:41,100 --> 00:02:41,466
All right.

61
00:02:41,466 --> 00:02:42,300
So there we go.

62
00:02:42,300 --> 00:02:46,533
Let's call first the PLT module pyplot.

63
00:02:46,933 --> 00:02:50,233
Then from this module
we're going to call the plot function.

64
00:02:50,566 --> 00:02:51,666
There you go.

65
00:02:51,666 --> 00:02:54,633
And in this function will same
as in the scatter function

66
00:02:54,633 --> 00:02:59,266
we have to input the coordinates
of the points or line nodes

67
00:02:59,266 --> 00:03:02,100
corresponding to the predictions.
So let's do this according to you.

68
00:03:02,100 --> 00:03:04,866
What is the first coordinate
of these predicted salaries.

69
00:03:04,866 --> 00:03:08,266
Well that's of course X train Y is.

70
00:03:08,266 --> 00:03:11,933
That is because right now we're
visualizing the result of the training set

71
00:03:12,100 --> 00:03:12,566
okay.

72
00:03:12,566 --> 00:03:13,733
So these corresponds

73
00:03:13,733 --> 00:03:17,700
to the numbers of years of experience
of the employees in the training set.

74
00:03:17,700 --> 00:03:19,900
That's the X coordinates okay.

75
00:03:19,900 --> 00:03:22,600
Now we're going to input
the y coordinates.

76
00:03:22,600 --> 00:03:24,666
And according to you which ones are these.

77
00:03:24,666 --> 00:03:27,000
Well it's something
we haven't created yet.

78
00:03:27,000 --> 00:03:30,066
You know we created actually the y print

79
00:03:30,066 --> 00:03:33,066
variable containing the predicted
salaries of the test set.

80
00:03:33,133 --> 00:03:36,800
But we haven't created a vector
containing the predicted salaries

81
00:03:36,800 --> 00:03:39,000
of the training set.
And that's totally fine.

82
00:03:39,000 --> 00:03:43,333
What we'll actually enter here
as the y coordinate of this plot

83
00:03:43,333 --> 00:03:44,433
we're about to make,

84
00:03:44,433 --> 00:03:48,066
are going to be exactly the predicted
salaries of the training set.

85
00:03:48,066 --> 00:03:53,733
And to get them I'm going to copy this
and paste it right here.

86
00:03:53,733 --> 00:03:56,733
But instead of inputting here X test

87
00:03:56,733 --> 00:04:02,200
I'm going to input actually X train
because calling the predict method

88
00:04:02,200 --> 00:04:05,500
on X train,
meaning on the numbers of years

89
00:04:05,500 --> 00:04:08,766
of experience of the employees
in the training set will get me

90
00:04:08,766 --> 00:04:11,933
exactly the predicted
salaries of the training set.

91
00:04:12,100 --> 00:04:15,100
All right,
just as before with predict X test.

92
00:04:15,433 --> 00:04:17,466
So xtrain. Yes.

93
00:04:17,466 --> 00:04:20,666
And then just as before,
we're going to choose a color

94
00:04:20,766 --> 00:04:24,333
which this time will be blue, right.

95
00:04:24,500 --> 00:04:26,100
So that we can clearly see the difference

96
00:04:26,100 --> 00:04:29,100
between the real salaries
and the predicted salaries.

97
00:04:29,100 --> 00:04:31,800
Okay.
So this will plot the regression line.

98
00:04:31,800 --> 00:04:33,400
Now next step what can we do.

99
00:04:33,400 --> 00:04:35,766
So here we're going to add a title.

100
00:04:35,766 --> 00:04:39,200
And to do this
we have to call the title function.

101
00:04:39,600 --> 00:04:42,833
And inside in the parentheses
we just have to input in quotes.

102
00:04:43,033 --> 00:04:45,133
Well the title we want for our graph.

103
00:04:45,133 --> 00:04:47,733
And so let's say salary

104
00:04:47,733 --> 00:04:50,566
versus experience okay.

105
00:04:50,566 --> 00:04:53,566
You can choose any other title
if you prefer another one.

106
00:04:53,900 --> 00:04:55,200
So salary versus experience.

107
00:04:55,200 --> 00:04:56,533
And then we're going to specify

108
00:04:56,533 --> 00:05:00,700
the training set because indeed
then we'll do the same for the test set.

109
00:05:01,433 --> 00:05:05,966
Now we're just going to add a label
to the x axis and to the y axis.

110
00:05:05,966 --> 00:05:09,500
And for this
we have to use the x label function

111
00:05:10,100 --> 00:05:14,966
into which we have to input well the label
we want to display on the x axis.

112
00:05:15,300 --> 00:05:20,400
And we're going to choose and quote
of course years of experience.

113
00:05:21,466 --> 00:05:22,200
All right.

114
00:05:22,200 --> 00:05:23,366
Good.

115
00:05:23,366 --> 00:05:28,433
Now we're going to copy this because
we're going to do the same for the y axis.

116
00:05:28,766 --> 00:05:32,033
And the name of the function for
this is y label.

117
00:05:32,400 --> 00:05:33,600
And for the y label.

118
00:05:33,600 --> 00:05:37,800
Well we will choose to display salary
okay.

119
00:05:38,233 --> 00:05:43,333
And finally in order to show the graphic
well we just had to finish here

120
00:05:43,333 --> 00:05:46,533
with PLT dot show,
you know the show function.

121
00:05:46,866 --> 00:05:49,866
And this will display
the graphic in the output.
