1
00:00:00,560 --> 00:00:03,980
In the last video we created this bad boy here.

2
00:00:04,010 --> 00:00:09,620
Now let's try and make something similar but using our own data so we can practice a little bit of data

3
00:00:09,620 --> 00:00:15,290
manipulation along with plotting and put some of our pandas in map plot lib skills together.

4
00:00:15,770 --> 00:00:21,680
So what we'll do we'll format the price column first we'll get it into integer format because right

5
00:00:21,680 --> 00:00:26,870
now we don't want those dollar signs and columns and commas we don't want those for when we're doing

6
00:00:26,870 --> 00:00:28,700
numerical operations.

7
00:00:28,700 --> 00:00:33,750
So to do that we'll go car sales and we'll change the price column.

8
00:00:33,770 --> 00:00:37,540
We need to access the string we saw this in the pandas section.

9
00:00:37,670 --> 00:00:39,380
String dot replace.

10
00:00:39,380 --> 00:00:43,140
And now we want to pass it a little bit of regex here.

11
00:00:43,220 --> 00:00:44,180
There we go.

12
00:00:44,180 --> 00:00:50,420
So essentially all this is saying is get the dollar sign the comma and the DOT which is here and here

13
00:00:50,750 --> 00:00:57,300
and replace it and we're going to replace it with an empty string.

14
00:00:57,380 --> 00:00:59,180
So we'll see what this does.

15
00:00:59,180 --> 00:01:06,760
Car sales we need to add car sales there so let's see what it looks like now.

16
00:01:06,780 --> 00:01:07,170
OK.

17
00:01:07,200 --> 00:01:07,680
Beautiful.

18
00:01:07,740 --> 00:01:10,950
But we've got two extra zeros on the end so we don't want that.

19
00:01:11,100 --> 00:01:16,060
So we want to go remove last two zeros.

20
00:01:16,080 --> 00:01:20,860
We've got car sales want to access the price column again.

21
00:01:20,860 --> 00:01:26,120
We're going to reassign it to car sales price.

22
00:01:26,270 --> 00:01:27,870
Again it's still a string.

23
00:01:28,280 --> 00:01:38,710
So if we have a look at this guy here type just to exemplify what we're doing with the DOT string it's

24
00:01:38,770 --> 00:01:40,300
a series.

25
00:01:40,310 --> 00:01:40,850
OK.

26
00:01:40,980 --> 00:01:41,620
But that's right.

27
00:01:41,620 --> 00:01:42,790
Well this is still a string.

28
00:01:42,910 --> 00:01:46,260
So that's we're accessing it with the DOT string method here.

29
00:01:46,300 --> 00:01:50,410
Maybe if we go 0 string there we go.

30
00:01:50,590 --> 00:01:54,430
So that's accessing this little one here and it's telling us that it's a string.

31
00:01:54,460 --> 00:01:58,540
So that's why we have to access it as a string first with DOT SDR.

32
00:01:59,140 --> 00:02:05,600
And now we want to remove the last two using slicing.

33
00:02:05,770 --> 00:02:09,570
Let's have a look at what car sales looks like here.

34
00:02:09,770 --> 00:02:10,620
Oh no.

35
00:02:10,790 --> 00:02:12,230
We've removed too many.

36
00:02:12,240 --> 00:02:13,690
So what we can do is rerun it.

37
00:02:13,700 --> 00:02:20,470
We're gonna have to re import our data frame to fix that little issue that's going to just reset it.

38
00:02:20,690 --> 00:02:26,180
Because remember Jupiter sales are running in sequential mode so if we run this sell more than once

39
00:02:26,420 --> 00:02:32,900
it's going to continually remove the last two digits of our price column as long as we keep running

40
00:02:32,900 --> 00:02:33,080
this.

41
00:02:33,080 --> 00:02:33,920
So we don't want that.

42
00:02:33,950 --> 00:02:42,590
So now our price column is in the format that we want we're going to add a sale date column which will

43
00:02:42,590 --> 00:02:54,790
be similar to the PD date range function that we used up here paid a date range so let's do that data

44
00:02:54,790 --> 00:02:55,180
range.

45
00:02:55,220 --> 00:02:56,580
1 1.

46
00:02:56,650 --> 00:03:00,910
We want to get past the date 2020 number of periods.

47
00:03:00,980 --> 00:03:04,490
So this is how many dates we want to add.

48
00:03:04,490 --> 00:03:06,790
We've got periods here.

49
00:03:06,890 --> 00:03:08,810
So that's a parameter.

50
00:03:08,810 --> 00:03:10,650
So that's going to add this one.

51
00:03:10,670 --> 00:03:11,870
We had a thousand periods.

52
00:03:11,870 --> 00:03:14,780
What we don't want that many for our car sales data frame.

53
00:03:14,870 --> 00:03:18,440
We actually just want date period for each row here.

54
00:03:18,620 --> 00:03:27,280
So we can do that by going periods Eagles Lane car sales and then we'll check our car sales data frame

55
00:03:27,280 --> 00:03:27,640
again.

56
00:03:28,760 --> 00:03:29,480
Beautiful.

57
00:03:29,480 --> 00:03:32,260
So now we've got a sale date here which is sequential.

58
00:03:32,300 --> 00:03:32,810
That's okay.

59
00:03:32,810 --> 00:03:37,970
Maybe they all weren't sold on sequential dates but we're just making an example of how we can start

60
00:03:37,970 --> 00:03:40,910
to manipulate our existing data frames.

61
00:03:41,000 --> 00:03:45,630
So the next thing what we want to do is we want to make a total column.

62
00:03:45,680 --> 00:03:49,970
So we want to start to add up how much we've made from selling all our cars.

63
00:03:49,970 --> 00:03:53,040
So if we added 4000 to 5000 to 7000.

64
00:03:53,120 --> 00:04:02,480
So this is where we can use something like Come some because it's going to add each of these prices

65
00:04:02,990 --> 00:04:07,100
cumulatively so let's make a total column.

66
00:04:07,210 --> 00:04:09,130
This should be car sales.

67
00:04:09,250 --> 00:04:14,170
Total sales and then what do we want we want the price column.

68
00:04:14,170 --> 00:04:14,700
That's right.

69
00:04:14,700 --> 00:04:17,500
Car sales rise.

70
00:04:17,800 --> 00:04:19,300
Dot com some.

71
00:04:19,300 --> 00:04:22,030
Let's see what happens here.

72
00:04:22,060 --> 00:04:27,350
Car sales Oh what have we done.

73
00:04:27,410 --> 00:04:28,600
It's worked.

74
00:04:28,820 --> 00:04:35,720
But because our price column is still in a string format it's just concatenated the strings to each

75
00:04:35,720 --> 00:04:36,380
other.

76
00:04:36,410 --> 00:04:37,280
So let's have a look.

77
00:04:37,280 --> 00:04:46,830
Car sales price zero daytime time Oh that's a bit boring.

78
00:04:47,130 --> 00:04:52,050
Let's get it back to exemplifying that it's a string.

79
00:04:52,050 --> 00:04:52,800
There we go.

80
00:04:52,800 --> 00:04:54,990
So our price column is still a string.

81
00:04:54,990 --> 00:05:03,120
So to fix this we have to change it first into an integer and then call the come some function so we

82
00:05:03,120 --> 00:05:09,440
can do dot as type int dot come some.

83
00:05:09,440 --> 00:05:17,400
So what this is going to do is go reassign the total sales column to be the price column as an integer

84
00:05:17,550 --> 00:05:22,230
as type int dot cumulative sum.

85
00:05:22,250 --> 00:05:29,230
This should work in theory but as always run the code then check Beautiful.

86
00:05:29,260 --> 00:05:29,710
OK.

87
00:05:29,740 --> 00:05:36,310
So this has gone total sales of thousand plus five thousand is nine thousand plus seven thousand is

88
00:05:36,310 --> 00:05:41,650
sixteen thousand and then it keeps going right up to the end so after we've sold all of our cars across

89
00:05:41,650 --> 00:05:46,570
these 10 days or so we have seventy six thousand four hundred and fifty.

90
00:05:46,570 --> 00:05:47,590
Beautiful.

91
00:05:47,650 --> 00:05:48,770
That's some good profit there.

92
00:05:48,810 --> 00:05:50,230
You can use that to fund.

93
00:05:50,470 --> 00:05:51,860
About a store.

94
00:05:51,910 --> 00:06:03,160
And now let's plot the total sales so we want car sales we want to see what our total sales looks like

95
00:06:03,190 --> 00:06:04,180
over time.

96
00:06:04,390 --> 00:06:09,480
So kind of replicating this graph here with our own data.

97
00:06:09,790 --> 00:06:20,200
Car sales dot plot x equals sale date and now this way you can cold up plot directly on a painter's

98
00:06:20,200 --> 00:06:26,890
data frame this is Panda's version of the map partly API or the wrapper Panda's wrapper for the map

99
00:06:26,890 --> 00:06:28,540
partly of API.

100
00:06:28,540 --> 00:06:29,370
So we're just going.

101
00:06:29,380 --> 00:06:35,410
Data frame dot plot and we're passing it out X which is the date and now y which is the total sales

102
00:06:35,410 --> 00:06:35,980
column.

103
00:06:36,450 --> 00:06:38,590
So let's see what this looks like.

104
00:06:38,650 --> 00:06:40,370
Total sales.

105
00:06:40,520 --> 00:06:46,060
They'll put a semicolon at the end so it doesn't print out beautiful.

106
00:06:46,090 --> 00:06:46,750
There we go.

107
00:06:46,750 --> 00:06:55,340
So this is just an example of how you can take some code write from the documentation or some example

108
00:06:55,340 --> 00:07:02,910
that you found online and we've replicated it up here but then we've taken the principles from it or

109
00:07:02,910 --> 00:07:11,240
some of them and put them together with our own data and then created this little visualization here.

110
00:07:11,250 --> 00:07:12,700
So now we can visualize.

111
00:07:12,750 --> 00:07:17,200
Okay so say we sold our 10 cars over the first 10 days of January.

112
00:07:17,280 --> 00:07:22,680
Now our total sales go on up into the riot which is very pleasing for our boss because that's the type

113
00:07:22,680 --> 00:07:29,020
of graph I want to see profits up into the right and so hopefully you starting to get an idea of the

114
00:07:29,020 --> 00:07:33,610
kind of workflow that you can go about when you're working on your own projects or working on problems

115
00:07:33,610 --> 00:07:35,910
you don't necessarily know the answer to.

116
00:07:35,920 --> 00:07:43,180
It's all about searching up trying to find some kind of example taking information from that and then

117
00:07:43,360 --> 00:07:46,850
seeing if it works with your own problem.

118
00:07:46,870 --> 00:07:48,780
Now we've seen a line plot.

119
00:07:48,790 --> 00:07:52,560
How about we try scatter plot so we've got car sales.

120
00:07:52,600 --> 00:08:00,140
Plot X. Let's do it with the odometer this time a domino column.

121
00:08:00,430 --> 00:08:02,520
K M beautiful.

122
00:08:02,590 --> 00:08:03,800
Is that the right column name.

123
00:08:03,950 --> 00:08:05,810
Yep we've worked with this data frame.

124
00:08:05,870 --> 00:08:14,250
Also kind of familiar with what's happening price and then when you're calling the plot function there's

125
00:08:14,260 --> 00:08:23,140
a little parameter here called kind which we can feed the type scatter to hopefully get a scatter plot

126
00:08:23,140 --> 00:08:23,460
here.

127
00:08:25,070 --> 00:08:32,120
What's happening scatter requires y column to be numeric is our price column not numeric.

128
00:08:32,390 --> 00:08:35,290
It's still not numeric because here we've called as type int.

129
00:08:35,300 --> 00:08:42,040
So this is only going to turn the price column into an integer for running this single line of code.

130
00:08:42,050 --> 00:08:46,090
It's not going to reassign the price column to be an integer.

131
00:08:46,100 --> 00:08:49,880
These are still strings in the car sales data frame.

132
00:08:49,880 --> 00:09:00,210
So what we can do to make this work we want to reassign the price column to be an integer price an Eagles

133
00:09:01,110 --> 00:09:08,300
sales price as tied in because our prices are always going to be integers.

134
00:09:08,310 --> 00:09:15,900
Money is a numerical value so it's okay to reassign this data type to be integer they should work.

135
00:09:15,930 --> 00:09:30,570
So we got reassign the price column to int plot scatter plot with price column as numeric.

136
00:09:30,570 --> 00:09:36,060
So without this line of code we'll comment that out we're going to get an error here.

137
00:09:36,060 --> 00:09:41,970
Value era requires y column to be numeric but what we're doing with this line here we uncle meant it.

138
00:09:43,310 --> 00:09:48,420
To comment and on comment code we can do command Slash.

139
00:09:48,470 --> 00:09:50,770
So I'm just pressing command slash here.

140
00:09:50,900 --> 00:09:53,360
We're going to reassign it as a numeric column.

141
00:09:53,360 --> 00:09:55,930
So if we do this beautiful.

142
00:09:56,030 --> 00:10:00,560
Now we have a scatter plot of our odometer column versus price.

143
00:10:00,560 --> 00:10:07,750
And remember to remove this little piece of code here you can put a semicolon right at the end Excellent.

144
00:10:07,800 --> 00:10:13,500
Now I'm kind of seeing a few examples here of how we can start to plot directly from a panda's data

145
00:10:13,500 --> 00:10:14,100
frame.

146
00:10:14,100 --> 00:10:18,450
We'll see a few more but we'll end this video before it gets too long.

147
00:10:18,450 --> 00:10:23,790
So in the meantime before the next one maybe you'll try plot a couple of these columns together whatever

148
00:10:23,790 --> 00:10:25,800
kind of plot you want to do.

149
00:10:25,920 --> 00:10:26,760
You've seen a few.

150
00:10:26,760 --> 00:10:32,610
A bar graph of scatter plot a line plot to have a practice of that and I'll see you in the next video.