1
00:00:00,870 --> 00:00:02,700
- [Tutor] Up to this point in the videos,

2
00:00:02,700 --> 00:00:05,270
we've been focusing on a number

3
00:00:05,270 --> 00:00:08,570
of different built in
collections in Python.

4
00:00:08,570 --> 00:00:12,210
We've looked in detail
now at lists, tuples,

5
00:00:12,210 --> 00:00:15,230
and in this lesson we've been
focusing on dictionaries.

6
00:00:15,230 --> 00:00:18,810
Next up, we'd like to take a
look at the set collection,

7
00:00:18,810 --> 00:00:21,440
which is also built into Python.

8
00:00:21,440 --> 00:00:24,910
A set is a collection of unique values,

9
00:00:24,910 --> 00:00:28,700
an unordered collection of
unique values specifically,

10
00:00:28,700 --> 00:00:32,020
so you should never depend on the order

11
00:00:32,020 --> 00:00:34,640
in which items appear in a set.

12
00:00:34,640 --> 00:00:38,000
It is mainly used to determine membership,

13
00:00:38,000 --> 00:00:40,180
whether or not something is in a set

14
00:00:40,180 --> 00:00:43,230
and we're going to show you a
number of different operations

15
00:00:43,230 --> 00:00:46,230
that are used along those lines.

16
00:00:46,230 --> 00:00:49,540
So let's go ahead and
create our very first set.

17
00:00:49,540 --> 00:00:52,710
In this case, we have a set
of strings called colors.

18
00:00:52,710 --> 00:00:56,600
And you'll notice that sets
are delimited by curly braces,

19
00:00:56,600 --> 00:00:58,280
just like dictionaries.

20
00:00:58,280 --> 00:01:03,060
However, in a dictionary we
have a key, a colon and a value

21
00:01:03,060 --> 00:01:06,240
for each item in the comma separated list.

22
00:01:06,240 --> 00:01:10,290
Whereas in a set, we just
have a key in each item

23
00:01:10,290 --> 00:01:13,670
that we are passing into
the curly braces here.

24
00:01:13,670 --> 00:01:15,560
So another thing I wanna point out

25
00:01:15,560 --> 00:01:18,370
is that I don't have
all unique values here.

26
00:01:18,370 --> 00:01:21,680
So what happens is, as
the set gets created,

27
00:01:21,680 --> 00:01:25,320
each of these values is going
to be inserted into the set

28
00:01:25,320 --> 00:01:28,240
and then when we get to
the second copy of red,

29
00:01:28,240 --> 00:01:30,650
it will try to insert that into the set,

30
00:01:30,650 --> 00:01:32,320
see that the value is already there

31
00:01:32,320 --> 00:01:35,060
and simply discard the extra value.

32
00:01:35,060 --> 00:01:39,750
So duplicate elimination is
one of the key capabilities

33
00:01:39,750 --> 00:01:42,540
of sets it ignores the duplicates.

34
00:01:42,540 --> 00:01:45,900
So let's go ahead and evaluate colors

35
00:01:45,900 --> 00:01:47,800
and you'll notice two things about

36
00:01:47,800 --> 00:01:52,800
the resulting curly
bracketed list of keys.

37
00:01:52,920 --> 00:01:55,730
In this case, there's only one key

38
00:01:55,730 --> 00:01:58,580
with the value red even
though we started with two

39
00:01:58,580 --> 00:02:00,880
in the statement that we wrote up above.

40
00:02:00,880 --> 00:02:03,540
And you'll also notice
that the values are not

41
00:02:03,540 --> 00:02:06,950
in the same order that we specified

42
00:02:06,950 --> 00:02:09,940
when we created the
set in the first place.

43
00:02:09,940 --> 00:02:12,500
So you should never rely on the order

44
00:02:12,500 --> 00:02:17,100
in which the elements are
listed inside of a set.

45
00:02:17,100 --> 00:02:22,100
It is mainly again used for
membership style testing.

46
00:02:22,100 --> 00:02:24,500
Now you can determine how many elements,

47
00:02:24,500 --> 00:02:26,160
are in a set very easily.

48
00:02:26,160 --> 00:02:29,360
So we have that built in
len function once again

49
00:02:29,360 --> 00:02:31,580
and if you pass it a set,

50
00:02:31,580 --> 00:02:34,060
it will count up the number
of elements in that set

51
00:02:34,060 --> 00:02:35,770
and give that back to you.

52
00:02:35,770 --> 00:02:39,540
And you can also very easily
check whether a value,

53
00:02:39,540 --> 00:02:44,100
is in a set by using the
in and not in operators.

54
00:02:44,100 --> 00:02:47,320
So if you want to do a
membership test, for example,

55
00:02:47,320 --> 00:02:50,900
if you wanna see if red is in colors,

56
00:02:50,900 --> 00:02:53,260
you can write an expression like this,

57
00:02:53,260 --> 00:02:55,680
you might use that in an
if statement for example,

58
00:02:55,680 --> 00:02:57,930
and in this case, we can see that indeed,

59
00:02:57,930 --> 00:03:02,930
red is in colors whereas
if we do purple in colors,

60
00:03:04,070 --> 00:03:07,810
of course, that's false because
purple is not in colors.

61
00:03:07,810 --> 00:03:09,360
And along those same lines,

62
00:03:09,360 --> 00:03:13,690
if I then test not in
colors, purple not in colors,

63
00:03:13,690 --> 00:03:16,490
you can see that indeed, that's true.

64
00:03:16,490 --> 00:03:18,610
Now, as you might expect,

65
00:03:18,610 --> 00:03:21,820
you are able to iterate through sets,

66
00:03:21,820 --> 00:03:24,470
just like you can iterate through lists,

67
00:03:24,470 --> 00:03:27,340
and tuples, and dictionaries and strings.

68
00:03:27,340 --> 00:03:30,390
So, again, Python's really flexible

69
00:03:30,390 --> 00:03:32,450
in that you can do a
lot of the same things

70
00:03:32,450 --> 00:03:34,670
across many different data types.

71
00:03:34,670 --> 00:03:36,840
So sets are iterable

72
00:03:36,840 --> 00:03:38,640
and here's a little for loop,

73
00:03:38,640 --> 00:03:40,850
that's going to walk through each color

74
00:03:40,850 --> 00:03:43,480
in the set called colors and for each one,

75
00:03:43,480 --> 00:03:44,920
in this case, we're going to give you

76
00:03:44,920 --> 00:03:47,860
the all uppercase version of the string,

77
00:03:47,860 --> 00:03:50,320
separated from the next by a space.

78
00:03:50,320 --> 00:03:51,710
So let's go ahead and do that

79
00:03:51,710 --> 00:03:55,380
and you can see we get the
same values once again.

80
00:03:55,380 --> 00:03:59,140
However, I would like you to
notice a couple of things here.

81
00:03:59,140 --> 00:04:00,960
Number one, they're all uppercase,

82
00:04:00,960 --> 00:04:03,440
because we called upper on each element.

83
00:04:03,440 --> 00:04:06,210
And number two, they're
not in the same order

84
00:04:06,210 --> 00:04:08,690
that they were displayed up above either.

85
00:04:08,690 --> 00:04:10,920
So again, don't rely on the order

86
00:04:10,920 --> 00:04:13,140
of the elements within that set,

87
00:04:13,140 --> 00:04:16,020
I can't emphasize that enough.

88
00:04:16,020 --> 00:04:19,500
Now, you can also create sets,

89
00:04:19,500 --> 00:04:23,090
using a built in function called set.

90
00:04:23,090 --> 00:04:25,330
So to demonstrate this concept,

91
00:04:25,330 --> 00:04:28,700
let me first create a list called numbers.

92
00:04:28,700 --> 00:04:31,630
And in this particular
case, we have a list

93
00:04:32,860 --> 00:04:35,430
with the elements from zero to nine in it

94
00:04:35,430 --> 00:04:36,600
that we're going to create,

95
00:04:36,600 --> 00:04:38,350
and we're going to concatenate that

96
00:04:38,350 --> 00:04:41,160
with another list containing
zero through four.

97
00:04:41,160 --> 00:04:44,090
So there will be some
duplicates in this list.

98
00:04:44,090 --> 00:04:47,070
And to show you that we'll
display the list here.

99
00:04:47,070 --> 00:04:49,860
So you see zero through nine
and then zero through four.

100
00:04:49,860 --> 00:04:53,130
So zero through four are duplicated twice

101
00:04:53,130 --> 00:04:55,090
in this particular list.

102
00:04:55,090 --> 00:04:58,410
Now, when you use the
set built in function,

103
00:04:58,410 --> 00:05:01,000
and you give it a sequence of values,

104
00:05:01,000 --> 00:05:03,770
it's going to create a set object.

105
00:05:03,770 --> 00:05:06,340
And as you might expect,

106
00:05:06,340 --> 00:05:10,400
that set object will have
only the unique items

107
00:05:10,400 --> 00:05:12,720
from the argument sequence.

108
00:05:12,720 --> 00:05:14,890
So if I do that, you can see we wind up

109
00:05:14,890 --> 00:05:17,930
with a set containing
only zero through nine.

110
00:05:17,930 --> 00:05:20,410
And the duplicate values
at the end of the list

111
00:05:20,410 --> 00:05:24,320
were simply ignored as
the set was created.

112
00:05:24,320 --> 00:05:29,320
Now, you may recall that when
I created an empty dictionary,

113
00:05:29,890 --> 00:05:33,090
I did that with an empty
set of curly braces.

114
00:05:33,090 --> 00:05:37,110
Well, that means I can't use
an empty set of curly braces

115
00:05:37,110 --> 00:05:39,670
to define an empty set object.

116
00:05:39,670 --> 00:05:42,070
So if you do need an empty set,

117
00:05:42,070 --> 00:05:44,830
the way you're going to do
that is with the set function

118
00:05:44,830 --> 00:05:46,430
with no arguments,

119
00:05:46,430 --> 00:05:50,640
and that is going to be
represented in its string format,

120
00:05:50,640 --> 00:05:53,740
as the word set followed
by empty parentheses.

121
00:05:53,740 --> 00:05:56,770
So sets are a little bit different

122
00:05:56,770 --> 00:06:01,570
in terms of how they are
represented in string format

123
00:06:01,570 --> 00:06:04,150
for an empty set versus an empty list

124
00:06:04,150 --> 00:06:07,980
which would be displayed
as square brackets,

125
00:06:07,980 --> 00:06:11,390
an empty tuple, which would
be displayed as parentheses,

126
00:06:11,390 --> 00:06:13,560
or an empty dictionary,

127
00:06:13,560 --> 00:06:17,500
which would be displayed
as a set of curly braces.

128
00:06:17,500 --> 00:06:21,610
Now, just for your
information sets are mutable,

129
00:06:21,610 --> 00:06:24,690
meaning you can modify their contents.

130
00:06:24,690 --> 00:06:29,010
The elements within sets,
however, have to be immutable,

131
00:06:29,010 --> 00:06:32,170
because they have to also be hashable,

132
00:06:32,170 --> 00:06:35,850
only immutable objects can be hashable.

133
00:06:35,850 --> 00:06:40,390
And the hashability is what
enables items to be placed

134
00:06:40,390 --> 00:06:44,960
into a set or a dictionary
and allows the data structures

135
00:06:44,960 --> 00:06:47,930
to maintain those unique keys.

136
00:06:47,930 --> 00:06:51,520
Now, it turns out that there is a type

137
00:06:51,520 --> 00:06:55,420
in the Python standard
library called frozen set,

138
00:06:55,420 --> 00:06:58,740
which is an immutable
representation of a set.

139
00:06:58,740 --> 00:07:03,450
So if in fact you need a
frozen set of elements,

140
00:07:03,450 --> 00:07:05,610
meaning that once the set is created,

141
00:07:05,610 --> 00:07:07,520
it can never be modified.

142
00:07:07,520 --> 00:07:09,960
Then you can use the frozen set type

143
00:07:09,960 --> 00:07:13,030
and frozen sets because
they are immutable,

144
00:07:13,030 --> 00:07:16,170
if they also contain
only immutable objects

145
00:07:16,170 --> 00:07:18,170
which they have to for keys,

146
00:07:18,170 --> 00:07:22,430
you can then use them actually
as keys in other sets,

147
00:07:22,430 --> 00:07:24,893
or in dictionaries as well.