1
00:00:00,230 --> 00:00:01,470
Well welcome back.

2
00:00:01,560 --> 00:00:06,470
It's the last video we left off saying that this plot probably isn't ideal.

3
00:00:06,480 --> 00:00:13,320
We've got a few subplots going on here and we've been using the PI plot method plotting directly from

4
00:00:13,320 --> 00:00:17,760
a panda's data frame rather than the object orientated method.

5
00:00:17,760 --> 00:00:19,950
And I'm sure you're wondering you're going.

6
00:00:19,950 --> 00:00:24,720
Daniel you've introduced so many different ways of plotting and you're telling us there's two ways of

7
00:00:24,720 --> 00:00:25,350
doing it.

8
00:00:25,500 --> 00:00:28,330
And you're saying one is less ideal than the other.

9
00:00:28,560 --> 00:00:30,050
Which one should I use.

10
00:00:30,090 --> 00:00:31,920
And that's a perfectly great question.

11
00:00:31,920 --> 00:00:45,140
So which one should you use pi plot this map plot lib o method now.

12
00:00:45,650 --> 00:00:48,920
There is no real definitive answer here.

13
00:00:48,920 --> 00:00:53,860
The documentation on that plot level will push you towards the O method always.

14
00:00:53,900 --> 00:01:04,610
However when plotting something quickly okay to use the PI plot method.

15
00:01:04,610 --> 00:01:04,850
Right.

16
00:01:04,850 --> 00:01:09,110
So what we've seen up here when you want to get a quick visualization maybe not something like this

17
00:01:09,140 --> 00:01:11,100
but definitely something like this.

18
00:01:11,120 --> 00:01:15,600
It's okay to use the PI plot method that we've gone through in the last couple of videos.

19
00:01:15,980 --> 00:01:25,230
But when plotting something more advanced use the O method.

20
00:01:25,550 --> 00:01:28,020
So we're gonna see an example of that in this video.

21
00:01:28,190 --> 00:01:31,540
Something a little bit more advanced than just something singular like this.

22
00:01:31,550 --> 00:01:35,510
But if you're just making a quick plot like this you just wanna see the distribution of a column it's

23
00:01:35,520 --> 00:01:38,040
perfectly okay to go dot plot.

24
00:01:38,120 --> 00:01:41,320
Now we'll see the O method in action.

25
00:01:41,330 --> 00:01:45,480
Let's create a subset of our data frame so we can do a little bit of data analysis.

26
00:01:45,500 --> 00:01:49,630
Let's say we wanted to explore the patients which are over the age of 50.

27
00:01:49,640 --> 00:01:52,120
We only wanted to look at those patients.

28
00:01:52,200 --> 00:01:56,900
We want to create a little bit of a plot and start to do some data analysis on those.

29
00:01:56,960 --> 00:02:03,410
So we'll create another data frame called over 50 which is equal to our heart disease data frame which

30
00:02:03,410 --> 00:02:05,450
will get a little refresher here.

31
00:02:05,450 --> 00:02:06,990
So we know what it looks like.

32
00:02:07,010 --> 00:02:11,640
This is a kind of workflow you'll be going through is looking at your data frame and going okay.

33
00:02:11,720 --> 00:02:16,820
My colleague or my teammates and my boss want to know what some data analysis of our patients which

34
00:02:16,820 --> 00:02:18,450
are over 50.

35
00:02:18,450 --> 00:02:24,080
So let's do that we can access those by using boolean indexing.

36
00:02:24,150 --> 00:02:25,460
So we want heart disease.

37
00:02:25,500 --> 00:02:31,020
Actually we want to use the age column which are greater than 50.

38
00:02:31,020 --> 00:02:32,120
Wonderful.

39
00:02:32,490 --> 00:02:37,420
And then we want to have a look it out over 50 dataset.

40
00:02:37,550 --> 00:02:39,620
What have we got wrong here.

41
00:02:39,710 --> 00:02:43,310
Age Kiara I've used a capital I.

42
00:02:43,370 --> 00:02:46,220
That's right wonderful.

43
00:02:46,750 --> 00:02:51,330
And we might check the length of this 209.

44
00:02:51,340 --> 00:02:52,730
So we have two hundred and eight.

45
00:02:52,730 --> 00:02:56,920
This is about 300 or so of the original dataset.

46
00:02:56,920 --> 00:02:57,800
So that's fine.

47
00:02:57,820 --> 00:02:58,840
Go ahead.

48
00:02:58,840 --> 00:02:59,500
Wonderful.

49
00:02:59,500 --> 00:03:01,980
And again we've dropped out these two rows.

50
00:03:01,990 --> 00:03:05,710
So 37 41 using our boolean indexing.

51
00:03:05,710 --> 00:03:08,430
So now it goes 0 three four five six.

52
00:03:08,500 --> 00:03:15,070
But the rest of the columns are still the same How about we start with a scatter plot of The Age and

53
00:03:15,070 --> 00:03:20,030
cholesterol but we want to color it with the flavor of the target column.

54
00:03:20,050 --> 00:03:21,220
Well that's a lot going on there.

55
00:03:21,220 --> 00:03:24,130
Let's see the code first and then we'll figure it out from there.

56
00:03:24,160 --> 00:03:35,930
Over 50 dot plot kind equals scatter and we want X equals age and we want Y equals.

57
00:03:35,930 --> 00:03:43,040
We want the cholesterol column which is this one here C H O L is short for cholesterol and then we launch

58
00:03:43,040 --> 00:03:48,450
to cover it using the c So c is short for color by the target column.

59
00:03:48,560 --> 00:03:50,930
So we won't plot kind scatter.

60
00:03:50,930 --> 00:04:00,080
Yes that makes sense x age y call or child cholesterol kind of a hard word to say see agile and the

61
00:04:00,110 --> 00:04:01,310
color is target.

62
00:04:01,310 --> 00:04:04,970
Let's see what happens all right.

63
00:04:05,240 --> 00:04:07,010
So we can kind of see what's going on here.

64
00:04:07,010 --> 00:04:08,770
We've got Charles on the Y.

65
00:04:08,780 --> 00:04:09,550
That makes sense.

66
00:04:09,590 --> 00:04:11,270
X doesn't really show us much.

67
00:04:11,270 --> 00:04:19,680
And C is giving us this Calabar to show us that because target has values of 0 1 4 if someone does have

68
00:04:19,680 --> 00:04:21,470
heart disease or they don't.

69
00:04:21,480 --> 00:04:23,330
Does this look okay to you.

70
00:04:23,670 --> 00:04:24,960
To me it doesn't look very good.

71
00:04:24,990 --> 00:04:26,560
It's doing what we want it to do.

72
00:04:26,610 --> 00:04:28,050
It's got different dots on here.

73
00:04:28,380 --> 00:04:35,550
If their target value of 0 or 1 but it doesn't show us anything on the x axis and these colors aren't

74
00:04:35,550 --> 00:04:36,290
really that great.

75
00:04:36,300 --> 00:04:37,880
So let's spruce things up a little.

76
00:04:38,190 --> 00:04:41,080
And how can we do that with our method.

77
00:04:41,100 --> 00:04:47,160
So this is the pipeline method a.k.a. plotting directly from the panda's data frame with plot.

78
00:04:47,190 --> 00:04:49,760
Now let's do o method.

79
00:04:49,830 --> 00:04:55,940
We're going to recreate this but using the O method so let's go fig act.

80
00:04:55,960 --> 00:04:57,320
We've seen this before.

81
00:04:57,320 --> 00:05:05,180
BLT dot subplots and we might go fig size just to have some practice adjusting the size of our fingers

82
00:05:05,630 --> 00:05:07,410
and we'll go 10 6.

83
00:05:07,410 --> 00:05:10,070
We want a width of 10 and a height of six.

84
00:05:10,100 --> 00:05:14,680
These are in inches by the way I believe so over 50 dot plot.

85
00:05:15,200 --> 00:05:17,500
This is gonna look very similar to what we've done up here.

86
00:05:17,930 --> 00:05:22,920
But there's one slight difference kind Eagles scatter.

87
00:05:23,270 --> 00:05:24,130
Wonderful.

88
00:05:24,140 --> 00:05:28,010
Now he could copy and paste but we're not going to do that because we're in the habit of writing out

89
00:05:28,010 --> 00:05:32,270
code Y equals charcoal or coal.

90
00:05:33,290 --> 00:05:39,940
See we use the target column as hell off and then the last part is the ax parameter.

91
00:05:40,040 --> 00:05:47,660
So when we're using the O method with panders the plot function gets this ax parameter tell us hey we're

92
00:05:47,660 --> 00:05:55,790
using the O method and we want you to plot this data on this axis the one we've just created.

93
00:05:55,880 --> 00:06:02,880
So let's say that an action ax would put a little semicolon at the end beautiful.

94
00:06:02,900 --> 00:06:07,300
So what we've done is we've recreated the same plot but it's already looking a little bit better.

95
00:06:07,370 --> 00:06:12,350
It's a bigger because we've put the fig sized parameter in there and now because we've passed the ax

96
00:06:12,350 --> 00:06:18,410
method we've now got this age column which has been labelled beautiful actually we might give this one

97
00:06:18,410 --> 00:06:28,780
a chance and say hey let's up this fig size and see if it does the same thing 10 6 we still don't have

98
00:06:28,780 --> 00:06:31,150
the age X label there.

99
00:06:31,270 --> 00:06:34,610
Well the 0 method is already pulling out ahead.

100
00:06:34,630 --> 00:06:40,050
Now what we might do is we set the X limit to be a bit wider so we can have more space on our data frame

101
00:06:40,060 --> 00:06:45,300
so let's have a look that looks like we're using the 0 0 method we can access the axis here.

102
00:06:45,320 --> 00:06:50,800
Remember the Axis is just this space that we're plotting data on if we go back to the figure of our

103
00:06:51,700 --> 00:06:55,080
anatomy of a map plot lib figure or plot axis.

104
00:06:55,090 --> 00:06:57,520
Is this part here where we're adding data.

105
00:06:57,640 --> 00:07:01,390
Whereas figure is the entire thing.

106
00:07:01,420 --> 00:07:16,490
So let's go back we want X dot set X Lim to be 45 100 can we figure out what this does before we even

107
00:07:16,490 --> 00:07:16,880
look at it.

108
00:07:17,330 --> 00:07:22,160
So set X Lim We could do shift tab but I want you to just have a think about it.

109
00:07:22,160 --> 00:07:26,940
So Ax set X Lim 45 100.

110
00:07:27,110 --> 00:07:27,700
So that's right.

111
00:07:27,710 --> 00:07:30,730
If it didn't get it if you press shift tab and read it you cheat a little.

112
00:07:30,730 --> 00:07:31,630
That's also fine.

113
00:07:31,640 --> 00:07:36,560
That's actually a good thing to do is start exploring what Chef Tab does but this is going to change

114
00:07:36,560 --> 00:07:43,990
the limitations of our x axis to be from 45 to 100 instead of what it is now which is about 50 to 80.

115
00:07:44,030 --> 00:07:50,710
So let's see what happens so see how we've got like a lot more whitespace over here now because we've

116
00:07:50,830 --> 00:07:56,410
upped the X limit but if we didn't want to do that we could comment that out and that looks a bit better.

117
00:07:56,500 --> 00:08:01,900
Just given an example of how you could quickly adjust the limitations of your accesses if they don't

118
00:08:01,900 --> 00:08:02,710
look too good.

119
00:08:02,780 --> 00:08:09,210
But lib is generally possibility good at creating axis limits that are suited to whatever data that

120
00:08:09,250 --> 00:08:10,330
you're using.

121
00:08:10,380 --> 00:08:14,340
Now this plot is okay but it definitely could look nicer.

122
00:08:14,620 --> 00:08:19,850
So this black and white is kind of hard to get this color bar is not really offering much at all.

123
00:08:19,930 --> 00:08:25,470
So let's take a little break there and in the next video we'll make this plot look a little bit nicer

124
00:08:25,720 --> 00:08:27,640
and we'll build off what we've already done here.