1
00:00:00,450 --> 00:00:01,500
Welcome back.

2
00:00:01,500 --> 00:00:06,960
In the last lecture we looked at a few different ways to manipulate data especially missing data.

3
00:00:06,960 --> 00:00:08,640
We looked at in place.

4
00:00:08,640 --> 00:00:13,630
So you want to perform an operation on Panda's data frames in place you can set the pace parameters

5
00:00:13,630 --> 00:00:18,590
to true or you can reassign it by using.

6
00:00:18,590 --> 00:00:25,150
We don't have a reassignment example where you can reassignment by setting it to equaling something.

7
00:00:25,240 --> 00:00:28,230
Now we've had enough of working with missing data.

8
00:00:28,390 --> 00:00:34,810
How do we create data from existing data or pandas has a few different ways that you can create data

9
00:00:34,870 --> 00:00:36,800
such as building a new column.

10
00:00:36,940 --> 00:00:38,590
Let's have a look at a couple of different ways.

11
00:00:39,250 --> 00:00:42,690
Let's start with a series so communist first.

12
00:00:42,910 --> 00:00:52,130
We'll go a column from series then we'll create a series called seats column equals Payday series and

13
00:00:52,130 --> 00:00:59,250
maybe inside that series we'll put a few numbers of five because a few of our cars have five seats or

14
00:00:59,360 --> 00:01:02,510
actually all of our cars have five seats but we're not going to fill it out.

15
00:01:02,550 --> 00:01:06,770
So at car sales seats another way to make a new column.

16
00:01:06,920 --> 00:01:15,890
We'll a comment here new column called seat remember our car sales data frame is the one we were working

17
00:01:15,890 --> 00:01:17,160
with above.

18
00:01:17,870 --> 00:01:19,820
So we can view it here.

19
00:01:19,890 --> 00:01:22,070
So car sales new column.

20
00:01:22,430 --> 00:01:27,160
Now it's very similar to how you would select a column to create a new column.

21
00:01:27,170 --> 00:01:30,360
But remember we're going to run the code first before we discuss it.

22
00:01:30,410 --> 00:01:38,320
So we're going to set it equal to the seats column which is that series we've just created and now we

23
00:01:38,320 --> 00:01:43,470
want to view our car sales data from excellent.

24
00:01:43,560 --> 00:01:46,610
So we'll see what this is done.

25
00:01:46,630 --> 00:01:52,590
We've got the exact same columns as before but this time we've now got seats on the right hand side.

26
00:01:52,680 --> 00:01:54,690
So that's created a column on the very right.

27
00:01:54,700 --> 00:01:59,860
When you create a new column in panders by default it appears on the very right hand side of your data

28
00:01:59,860 --> 00:02:05,040
frame but we've got some missing values here.

29
00:02:05,660 --> 00:02:06,560
Okay.

30
00:02:06,670 --> 00:02:14,350
That's because our series was only five by five so it's length five with the value five I'm getting

31
00:02:14,350 --> 00:02:16,480
confused all these fives here.

32
00:02:16,630 --> 00:02:22,810
So this is just to say this Toyota white car 400 thousand dollars because we haven't fixed up that price

33
00:02:22,810 --> 00:02:24,780
column and it has five seats.

34
00:02:24,880 --> 00:02:32,830
But how would we fill this in and maybe we can use car sales with the seats column we saw this before

35
00:02:32,830 --> 00:02:39,220
in the previous video fill in a we want five.

36
00:02:39,220 --> 00:02:39,700
There we go.

37
00:02:40,080 --> 00:02:41,140
All we want in place.

38
00:02:41,150 --> 00:02:41,370
Yeah.

39
00:02:41,380 --> 00:02:44,770
So we don't have to implies equals true shift.

40
00:02:44,770 --> 00:02:47,040
Now we have a look at car sales.

41
00:02:47,500 --> 00:02:49,160
Mistake.

42
00:02:49,240 --> 00:02:49,870
There we go.

43
00:02:49,900 --> 00:02:51,550
All of our cars have five seats.

44
00:02:51,550 --> 00:02:52,840
Beautiful.

45
00:02:52,840 --> 00:02:54,200
We kind of already knew that already.

46
00:02:54,220 --> 00:02:56,450
Well maybe some didn't have five seats.

47
00:02:56,470 --> 00:03:01,120
Now there's another way to create a column and that's from a point in the list which is similar to our

48
00:03:01,120 --> 00:03:10,330
series column from list might type in Python at least so we won't maybe fuel economy that's something

49
00:03:10,330 --> 00:03:16,120
else you want to consider when you buy a car these days you want to make sure that you're getting the

50
00:03:16,120 --> 00:03:18,610
best value or the best bang for your buck.

51
00:03:18,610 --> 00:03:19,810
Did I spell economy.

52
00:03:19,830 --> 00:03:20,390
No.

53
00:03:20,410 --> 00:03:21,400
Mix them up.

54
00:03:21,430 --> 00:03:22,260
There we go.

55
00:03:22,790 --> 00:03:30,600
We want maybe seven point five litres per 100 kilometres some nine point two 5.0 nine point six.

56
00:03:30,600 --> 00:03:32,840
We're just making these numbers up here.

57
00:03:32,990 --> 00:03:33,940
There we go.

58
00:03:33,940 --> 00:03:35,880
It's about five long as well.

59
00:03:35,890 --> 00:03:42,390
We want car sales what should we call this column fuel per hundred came.

60
00:03:42,420 --> 00:03:44,300
That's nice and easy to remember.

61
00:03:44,470 --> 00:03:45,130
Equals.

62
00:03:45,150 --> 00:03:52,290
We're going to set it to being equal to our fuel economy list economy and then we're going to have a

63
00:03:52,290 --> 00:03:56,580
look at it here value era.

64
00:03:56,590 --> 00:04:04,810
What do we get wrong length of value does not match length of index Huh says an interesting era.

65
00:04:04,900 --> 00:04:07,360
So maybe we need more values in here.

66
00:04:07,390 --> 00:04:09,310
Let's keep filling it up.

67
00:04:09,820 --> 00:04:11,230
Eight point seven three.

68
00:04:11,260 --> 00:04:12,850
This is a really fuel efficient car.

69
00:04:12,850 --> 00:04:13,900
See what happens.

70
00:04:13,900 --> 00:04:14,800
Still not the same.

71
00:04:14,800 --> 00:04:15,430
What do we need.

72
00:04:15,430 --> 00:04:16,150
Zero.

73
00:04:16,360 --> 00:04:23,630
One two three four five six seven eight mumble should do four or five.

74
00:04:23,770 --> 00:04:24,460
There we go.

75
00:04:25,110 --> 00:04:32,600
So if you want to create it from a list did that on purpose wink wink it has to be the same length as

76
00:04:32,600 --> 00:04:39,920
your existing data frame whereas a series can be different kind of length to what your data frame already

77
00:04:39,920 --> 00:04:40,480
is.

78
00:04:40,490 --> 00:04:47,030
So we needed nine or 10 values in here because it starts from zero because our original car sales data

79
00:04:47,030 --> 00:04:50,900
frame was of length 10 all right.

80
00:04:50,900 --> 00:04:52,390
Let's have another look at the way.

81
00:04:52,550 --> 00:04:56,270
What if we wanted to create a column from another column.

82
00:04:56,270 --> 00:05:03,040
So as if they wanted to figure out how much should this car has fuel it's used in its lifetime.

83
00:05:03,590 --> 00:05:09,800
So if this is per 100 kilometers it's done a hundred and fifty thousand kilometers.

84
00:05:09,810 --> 00:05:16,950
Maybe if we divided the odometer column by one hundred and times it by this column we could work out

85
00:05:16,950 --> 00:05:21,970
how many litres of fuel this car is used in its whole entire life.

86
00:05:21,990 --> 00:05:23,770
Let's try that out.

87
00:05:23,790 --> 00:05:26,430
So if we go car sales door.

88
00:05:26,490 --> 00:05:27,090
What's a good name.

89
00:05:27,090 --> 00:05:32,040
Total fuel used beautiful equals car sales.

90
00:05:32,040 --> 00:05:39,920
We want the motor column car sales odometer maybe we press tab area tab auto complete.

91
00:05:40,370 --> 00:05:46,850
Now we want to divide this by a hundred because this is fuel per 100 kilometers and so we could times

92
00:05:46,850 --> 00:05:49,460
that by car sales.

93
00:05:49,650 --> 00:05:53,880
We want fuel her tab per tab.

94
00:05:53,880 --> 00:05:57,180
No I press tab to get fuel per 100 km.

95
00:05:57,260 --> 00:06:03,380
Now what do you think this will do this is gonna create a new column because this string doesn't exist

96
00:06:03,380 --> 00:06:08,300
in our data frame that's going to take the odometer column another beautiful thing about Pan is numeric

97
00:06:08,300 --> 00:06:14,960
columns is that you can perform operations directly on them and then it's going to divide the odometer

98
00:06:14,960 --> 00:06:23,350
column by one hundred and times it by the fuel per 100 km column let's see what happens we want to view

99
00:06:23,350 --> 00:06:31,500
our data from beautiful so this is gonna be total fuel used in liters or maybe we could have known that

100
00:06:31,500 --> 00:06:41,240
better if we create this new column by typing in liters l it's gonna create another column total fuel

101
00:06:41,240 --> 00:06:49,940
used l That's right we don't mind that for now so 11 thousand two hundred eighty three ladies if you

102
00:06:49,940 --> 00:06:56,150
did that and one guy that'd be quite expensive depending on the price of fuel where you live right now

103
00:06:56,570 --> 00:07:05,240
is one more easier way to create a column we can do it from create a column from a single value let's

104
00:07:05,240 --> 00:07:05,750
do that.

105
00:07:05,750 --> 00:07:08,370
Car sales maybe we want a simple one.

106
00:07:08,480 --> 00:07:15,810
Number of wheels equals 4 because if your car didn't have four wheels you really wouldn't be out of

107
00:07:15,810 --> 00:07:22,910
drive now and you can start to see how our data frame can evolve over time Ryan we have manipulated

108
00:07:22,950 --> 00:07:23,880
a few things.

109
00:07:23,880 --> 00:07:28,500
Maybe some of these columns aren't very useful in the long run but it's just an example of how you can

110
00:07:28,500 --> 00:07:30,960
quickly use panders to change up your data frame.

111
00:07:30,960 --> 00:07:35,670
We've adjusted the price column with lowered the make column we've added a bunch of our own columns

112
00:07:36,080 --> 00:07:40,250
is a step you'll take in a data size or machine learning project right.

113
00:07:40,290 --> 00:07:43,350
Maybe we'll add one more column.

114
00:07:43,560 --> 00:07:48,480
We want to know if our cars pass to road safety and by default it equals true.

115
00:07:48,480 --> 00:07:55,260
This is just to give another example of different types of columns D types.

116
00:07:55,350 --> 00:07:57,070
So yeah there we go.

117
00:07:57,240 --> 00:08:01,440
We've got a bull column on the end now so bull.

118
00:08:01,440 --> 00:08:02,360
True or false.

119
00:08:02,370 --> 00:08:06,630
We've got a few ints we've got a few floats we've got some object columns we've got true.

120
00:08:06,630 --> 00:08:13,250
So this is how versatile panders is right you can use many different data types within and now what

121
00:08:13,250 --> 00:08:17,670
if we wanted to remove one of our columns like this one because it's got the L at the top.

122
00:08:17,690 --> 00:08:19,370
That's less confusing than this.

123
00:08:19,370 --> 00:08:21,740
We don't know what measurement that's in.

124
00:08:21,770 --> 00:08:30,580
You can use the function drop it's going to drop a drop column so we'll go total if you will use.

125
00:08:30,580 --> 00:08:32,140
That's our column name.

126
00:08:32,140 --> 00:08:36,220
Now we need to give it drop requires the parameter axis.

127
00:08:36,220 --> 00:08:39,780
So axis equals one if you remember the anatomy of our data frame.

128
00:08:39,780 --> 00:08:41,440
Let's see if we have it here.

129
00:08:41,500 --> 00:08:45,420
That's because columns are on axis equals 1.

130
00:08:45,430 --> 00:08:47,680
Remember how I said some functions require access.

131
00:08:47,680 --> 00:08:52,420
It took me a while to remember this but that's what you have to remember is if you're talking about

132
00:08:52,420 --> 00:08:55,530
a column you're talking about access 1.

133
00:08:55,700 --> 00:08:57,990
Let's do that drop.

134
00:08:58,180 --> 00:09:00,410
Beautiful car sales.

135
00:09:00,970 --> 00:09:06,080
It's still got that we keep forgetting to re-emphasize this point.

136
00:09:06,100 --> 00:09:10,060
We need to either reassign or use in place method.

137
00:09:10,180 --> 00:09:16,660
So we're reassigning our car sales data frame here drop car sales.

138
00:09:16,660 --> 00:09:20,460
Now we've dropped in beautiful where that data frame.

139
00:09:20,530 --> 00:09:21,890
That's amazing.

140
00:09:21,910 --> 00:09:22,330
All right.

141
00:09:22,510 --> 00:09:25,450
So we've seen out to create a few different columns here.

142
00:09:25,450 --> 00:09:28,220
We've seen how to drop a column if we don't want it.

143
00:09:28,300 --> 00:09:31,350
There's a few more things we want to do for manipulating our data frames.

144
00:09:31,360 --> 00:09:35,220
But before this lecture gets too long again go back.

145
00:09:35,260 --> 00:09:38,130
I want you to if you're going to take a little break take a little break.

146
00:09:38,140 --> 00:09:42,080
But otherwise before the next election try creating a column of your own.

147
00:09:42,130 --> 00:09:46,750
Something can be as simple as you like but just practice typing out some code like this.

148
00:09:46,750 --> 00:09:52,390
Maybe you do a little operation can be as fancy or as UN fancy as you like but just practice making

149
00:09:52,380 --> 00:09:53,770
a new column of your own.

150
00:09:53,890 --> 00:09:56,110
And I'll see you back in the next video.