1
00:00:00,000 --> 00:00:01,155
[No audio]

2
00:00:01,180 --> 00:00:03,600
Good, we're continuing our series of

3
00:00:03,629 --> 00:00:06,387
lectures about manipulating Pandas dataframes.

4
00:00:06,678 --> 00:00:08,460
So you'll learn how to drop a

5
00:00:08,460 --> 00:00:10,740
row or column out of the DataFrame in

6
00:00:10,740 --> 00:00:12,750
the previous lecture. Now let's see how

7
00:00:12,750 --> 00:00:16,655
we can add a column or row to a DataFrame.

8
00:00:17,526 --> 00:00:19,146
So we have df7 here with

9
00:00:19,175 --> 00:00:22,775
Address as index column and the series

10
00:00:22,800 --> 00:00:26,070
of column labels. Now, just to clarify

11
00:00:26,070 --> 00:00:28,110
something, I know this is difficult to

12
00:00:28,110 --> 00:00:30,300
wrap up your mind, how you'll be using

13
00:00:30,300 --> 00:00:32,940
these operations in real life? So I just

14
00:00:32,940 --> 00:00:35,370
want you to know the syntax of doing

15
00:00:35,370 --> 00:00:37,860
this operations, and throughout the

16
00:00:37,860 --> 00:00:40,170
course, you will learn how to actually

17
00:00:40,200 --> 00:00:42,510
put this into work with real life

18
00:00:42,510 --> 00:00:44,400
examples. So don't worry about that.

19
00:00:45,035 --> 00:00:47,495
Anyway, let's add a colon there, and to do

20
00:00:47,520 --> 00:00:49,860
that, you'd want to say df and in

21
00:00:49,860 --> 00:00:51,960
square brackets, you pass the name of

22
00:00:51,960 --> 00:00:55,920
the new column, and let's say, Continent,

23
00:00:56,615 --> 00:00:58,865
that'd be equal to, you know, now you

24
00:00:58,890 --> 00:01:01,320
would have to pass a list of values that

25
00:01:01,320 --> 00:01:03,330
you want to populate that column with,

26
00:01:04,290 --> 00:01:08,430
let's say North America. Now, if you

27
00:01:08,455 --> 00:01:10,285
execute that, you'll get an error,

28
00:01:10,590 --> 00:01:12,330
name df is not defined. Actually, I

29
00:01:12,330 --> 00:01:15,180
didn't mean this error. This was another

30
00:01:15,180 --> 00:01:17,940
thing. You get this error. So it says

31
00:01:17,940 --> 00:01:20,520
that the length of values does not match

32
00:01:20,520 --> 00:01:24,690
the length of index. So the length of

33
00:01:24,720 --> 00:01:28,800
index is, you know, the length of index is

34
00:01:29,730 --> 00:01:33,390
df7.index, the length of your

35
00:01:33,390 --> 00:01:36,000
index is 5, but you're trying to pass

36
00:01:36,000 --> 00:01:39,420
their list with a length of 1. So

37
00:01:39,420 --> 00:01:41,970
you've got only one element there. So

38
00:01:41,970 --> 00:01:44,040
the solution here is to pass a list with

39
00:01:44,040 --> 00:01:47,310
the exact number of items that you have

40
00:01:47,310 --> 00:01:50,190
in your table, in your DataFrame. So we

41
00:01:50,190 --> 00:01:53,340
have 5 rows there, 5, and yeah,

42
00:01:53,340 --> 00:01:56,580
you could add four more items here,

43
00:01:56,580 --> 00:01:58,950
North America, North America, etc, or

44
00:01:58,950 --> 00:02:00,660
you can do some fancy things in there.

45
00:02:00,900 --> 00:02:05,460
So you could say df7.shape then

46
00:02:05,460 --> 00:02:11,790
0 times that, execute, df7, print that

47
00:02:11,790 --> 00:02:15,390
out. Here, we have a new column. So if

48
00:02:15,390 --> 00:02:16,882
you're confused with this shape

49
00:02:16,907 --> 00:02:20,280
0 times North America, well, what I

50
00:02:20,280 --> 00:02:23,130
did is, you know, df7.shape,

51
00:02:23,160 --> 00:02:26,070
what you get is 5, 7, which means

52
00:02:26,070 --> 00:02:28,200
you have 5 rows and 7 columns.

53
00:02:29,580 --> 00:02:33,270
Now, what I want to get is the first

54
00:02:33,300 --> 00:02:36,420
item of that tuple. So I get 5 by

55
00:02:36,420 --> 00:02:39,300
doing shape 0. With this, I always

56
00:02:39,300 --> 00:02:41,880
make sure that I'm getting the number of

57
00:02:41,880 --> 00:02:45,180
rows that my DataFrame has, and then

58
00:02:45,180 --> 00:02:50,040
when you multiply 5 by a list with

59
00:02:50,310 --> 00:02:52,530
one element, you get a list with 5

60
00:02:52,555 --> 00:02:56,185
elements. America, that's the idea,

61
00:02:56,910 --> 00:02:59,730
delete that, and note that this is

62
00:02:59,730 --> 00:03:03,210
actually an inplace operation. So that

63
00:03:03,210 --> 00:03:05,400
will update your DataFrame, and all

64
00:03:05,400 --> 00:03:07,470
that it was about adding a new column, and

65
00:03:07,470 --> 00:03:08,915
how about modifying a new column.

66
00:03:10,595 --> 00:03:14,075
So modifying the Continent column, you

67
00:03:14,100 --> 00:03:15,960
know, that could be something like

68
00:03:16,950 --> 00:03:20,910
Country, and you can also add some

69
00:03:20,935 --> 00:03:24,391
streams in there, let's say plus a comma,

70
00:03:25,352 --> 00:03:29,380
plus, maybe another string, North America.

71
00:03:29,404 --> 00:03:32,126
[No audio]

72
00:03:32,151 --> 00:03:35,010
Execute that, print all

73
00:03:35,010 --> 00:03:39,120
the DataFrame, and yeah, see what we

74
00:03:39,120 --> 00:03:42,150
got. So what we did here is we updated

75
00:03:42,180 --> 00:03:44,820
the Continent data column by referring

76
00:03:44,850 --> 00:03:46,470
to an existing column, which is

77
00:03:46,470 --> 00:03:49,920
Country, and then for each value of the

78
00:03:49,920 --> 00:03:52,200
Country column, we added so we

79
00:03:52,200 --> 00:03:54,990
concatenate the comma string, which you

80
00:03:54,990 --> 00:03:57,900
can see in here, just after USA, and we

81
00:03:57,900 --> 00:04:01,650
also added another string, so North

82
00:04:01,650 --> 00:04:04,260
America. Now this could also be another

83
00:04:04,260 --> 00:04:07,350
column if you liked. So if you pass

84
00:04:07,350 --> 00:04:10,201
here, let's say Employees, you'd get

85
00:04:10,505 --> 00:04:12,305
8 instead of the first North

86
00:04:12,330 --> 00:04:16,950
America, and then 15, and so on. So that's

87
00:04:16,950 --> 00:04:20,010
how you update column. Now how about

88
00:04:20,040 --> 00:04:22,260
adding a new row? Well, this can be a

89
00:04:22,260 --> 00:04:24,630
bit tricky, but still understandable.

90
00:04:26,040 --> 00:04:28,200
What you could do here, because there's

91
00:04:28,200 --> 00:04:31,800
not easy methods to pass row to a

92
00:04:31,800 --> 00:04:34,980
DataFrame. What you could do is you

93
00:04:34,980 --> 00:04:38,340
could say df7_t, so I'm creating a

94
00:04:38,340 --> 00:04:43,388
new variable, that would be equal to df7.T.

95
00:04:43,713 --> 00:04:46,080
So T actually is a method

96
00:04:46,080 --> 00:04:48,900
that what it does is it transposes your

97
00:04:48,925 --> 00:04:52,534
DataFrame, and with transposition I mean,

98
00:04:52,714 --> 00:04:54,635
you know, df7, you check

99
00:04:54,660 --> 00:04:57,600
your new DataFrame. What you get is

100
00:04:57,600 --> 00:05:00,810
this. So you're rows have become

101
00:05:00,810 --> 00:05:03,330
columns, and your columns have become

102
00:05:03,330 --> 00:05:06,390
rows. So spend a few second looking at

103
00:05:06,390 --> 00:05:11,610
this, yep, and what we can do now is, you

104
00:05:11,610 --> 00:05:15,816
know, we can use the same syntax, df7_t,

105
00:05:16,049 --> 00:05:18,210
and then add a new column in there,

106
00:05:18,240 --> 00:05:20,460
so for the name of a column, you'll

107
00:05:20,460 --> 00:05:27,288
have to pass an Address, let's say My Address,

108
00:05:28,287 --> 00:05:30,180
let it be equal again to a

109
00:05:30,180 --> 00:05:33,810
list, and you want the list to reflect

110
00:05:34,571 --> 00:05:39,235
this order. So City first, let's say, My City,

111
00:05:40,438 --> 00:05:41,988
Country, My Country,

112
00:05:42,012 --> 00:05:44,536
[No audio]

113
00:05:44,581 --> 00:05:48,875
and 10 for Employees, and 7 for ID,

114
00:05:50,268 --> 00:05:54,487
My Shop for the shop name, My State,

115
00:05:54,511 --> 00:05:58,437
[No audio]

116
00:05:58,462 --> 00:05:59,618
and My Continent,

117
00:06:00,155 --> 00:06:06,545
execute that, and df7_t. If you look at

118
00:06:06,570 --> 00:06:08,813
these now, you'll see that

119
00:06:08,837 --> 00:06:10,837
[No audio]

120
00:06:10,862 --> 00:06:13,521
you get a new column in your DataFrame,

121
00:06:14,271 --> 00:06:20,919
and now what you do so to complete the trick is, let

122
00:06:21,000 --> 00:06:26,955
me say df7 equals to df7_t.T.

123
00:06:26,979 --> 00:06:29,723
[No audio]

124
00:06:29,748 --> 00:06:31,110
So df7 now,

125
00:06:31,260 --> 00:06:35,190
we'll have the new row added at the end. So

126
00:06:35,190 --> 00:06:37,410
let me wrap this up. Again, what I did

127
00:06:37,410 --> 00:06:40,110
is I transpose the original DataFrame,

128
00:06:40,140 --> 00:06:43,860
df7, then I added a column there.

129
00:06:44,040 --> 00:06:46,020
And now if I transpose the DataFrame

130
00:06:46,020 --> 00:06:48,000
back again to its original position,

131
00:06:48,360 --> 00:06:50,610
this column that I added will be

132
00:06:50,610 --> 00:06:54,300
converted to a row, and that does the

133
00:06:54,300 --> 00:06:58,020
trick, and similarly, you can modify a

134
00:06:58,020 --> 00:07:00,780
row if you like. So in that case, you

135
00:07:00,780 --> 00:07:02,880
don't point to My Address, but you'd

136
00:07:02,880 --> 00:07:06,630
point to an existing column with an

137
00:07:06,630 --> 00:07:08,550
address, with an existing address name.

138
00:07:08,765 --> 00:07:11,135
For instance, this one, if I pass it

139
00:07:11,160 --> 00:07:15,450
here, and execute that, execute that, and that,

140
00:07:17,015 --> 00:07:20,915
and you'll see that the values of this row

141
00:07:20,940 --> 00:07:23,610
with this address name have been updated

142
00:07:23,610 --> 00:07:27,360
now, as you can see in here. Now, that

143
00:07:27,360 --> 00:07:29,940
closes this lecture as well, and yeah,

144
00:07:29,940 --> 00:07:31,540
I'll see you later.