1
00:00:00,680 --> 00:00:01,940
- [Instructor] In the preceding video,

2
00:00:01,940 --> 00:00:04,360
we showed you a way to use a dictionary

3
00:00:04,360 --> 00:00:06,850
to summarize word counts.

4
00:00:06,850 --> 00:00:10,000
Now, when you're doing things like that,

5
00:00:10,000 --> 00:00:12,910
which are kind of common types of tasks,

6
00:00:12,910 --> 00:00:17,200
it's often the case that Python
already has a library class

7
00:00:17,200 --> 00:00:19,660
that can help you with that task.

8
00:00:19,660 --> 00:00:22,570
Summarizing word counts
is a common technique

9
00:00:22,570 --> 00:00:25,710
for text analysis and it turns out

10
00:00:25,710 --> 00:00:30,210
that there is a type in
the collections module,

11
00:00:30,210 --> 00:00:33,470
which is part of the
Python Standard Library

12
00:00:33,470 --> 00:00:37,293
that can help us perform that task easily.

13
00:00:38,344 --> 00:00:40,040
So far we've looked at a number

14
00:00:40,040 --> 00:00:41,810
of different built-in collections.

15
00:00:41,810 --> 00:00:45,550
We looked at lists, and
tuples, and now dictionaries,

16
00:00:45,550 --> 00:00:48,060
and soon, we're going
to look at sets as well.

17
00:00:48,060 --> 00:00:50,510
But there are other types
of collections out there

18
00:00:50,510 --> 00:00:53,660
and this particular module
actually has a bunch

19
00:00:53,660 --> 00:00:55,970
of additional collection types

20
00:00:55,970 --> 00:00:59,060
that you can use in your Python code.

21
00:00:59,060 --> 00:01:01,950
One of which is this Counter class.

22
00:01:01,950 --> 00:01:05,550
And, it's a type of
dictionary that knows how to,

23
00:01:05,550 --> 00:01:09,350
as they say, count hashable objects.

24
00:01:09,350 --> 00:01:12,110
Hashable objects are immutable objects

25
00:01:12,110 --> 00:01:15,670
that you can use as the
keys in a dictionary

26
00:01:15,670 --> 00:01:18,530
or that you can use as
the elements in a set,

27
00:01:18,530 --> 00:01:21,590
where the set elements must be unique.

28
00:01:21,590 --> 00:01:23,930
So anything hashable,
which would be things

29
00:01:23,930 --> 00:01:27,120
like strings and the numeric values,

30
00:01:27,120 --> 00:01:29,580
like integers and floating point numbers,

31
00:01:29,580 --> 00:01:33,810
typically they're going to be
immutable types of objects.

32
00:01:33,810 --> 00:01:36,150
Most commonly, they're
going to be strings.

33
00:01:36,150 --> 00:01:39,700
But we can take advantage of
this built-in Counter class

34
00:01:39,700 --> 00:01:43,410
in the collections module
to perform the work

35
00:01:43,410 --> 00:01:47,090
that we showed you in the
preceding script, in that loop.

36
00:01:47,090 --> 00:01:50,550
So it's going to do all
of that work for us.

37
00:01:50,550 --> 00:01:52,590
Now, to demonstrate this concept,

38
00:01:52,590 --> 00:01:56,030
let's switch over to the
terminal window here.

39
00:01:56,030 --> 00:02:00,140
And, let's go ahead and from
that collections module,

40
00:02:00,140 --> 00:02:03,200
let's import the Counter class.

41
00:02:03,200 --> 00:02:06,050
And, just to demonstrate, by the way,

42
00:02:06,050 --> 00:02:08,630
we're going to take a string,

43
00:02:08,630 --> 00:02:11,870
paste it in the same text statement

44
00:02:11,870 --> 00:02:14,040
that we created in the script.

45
00:02:14,040 --> 00:02:17,370
So we're going to once
again have this string,

46
00:02:17,370 --> 00:02:18,420
well actually two strings,

47
00:02:18,420 --> 00:02:21,540
that are going to be assembled
together into one string.

48
00:02:21,540 --> 00:02:24,080
And, remember what we
did in the script is,

49
00:02:24,080 --> 00:02:27,260
we told that string to split itself

50
00:02:27,260 --> 00:02:29,620
into a list of individual strings.

51
00:02:29,620 --> 00:02:31,450
So just to show you what that does,

52
00:02:31,450 --> 00:02:36,160
if I say text.split, you get a list.

53
00:02:36,160 --> 00:02:37,320
There's the square brackets

54
00:02:37,320 --> 00:02:39,370
denoting a list of string values,

55
00:02:39,370 --> 00:02:41,070
comment limited of course,

56
00:02:41,070 --> 00:02:43,100
and these are all the individual words

57
00:02:43,100 --> 00:02:46,860
that you see in the
text body up above here.

58
00:02:46,860 --> 00:02:47,980
So what we're going to do is

59
00:02:47,980 --> 00:02:52,410
feed text.split into the
creation of a Counter object

60
00:02:52,410 --> 00:02:54,490
and as part of creating the object,

61
00:02:54,490 --> 00:02:56,700
it's automatically going to summarize

62
00:02:56,700 --> 00:02:59,570
all of these words into their word counts.

63
00:02:59,570 --> 00:03:01,760
So to do that, we're
going to create an object,

64
00:03:01,760 --> 00:03:06,760
I'll call counter, with a lower
case c to differentiate it,

65
00:03:07,040 --> 00:03:08,220
from the class name.

66
00:03:08,220 --> 00:03:11,310
And, we'll say text.split here,

67
00:03:11,310 --> 00:03:14,640
and that's going to go
create the Counter object.

68
00:03:14,640 --> 00:03:17,230
And, to save me a tiny
bit of typing time here,

69
00:03:17,230 --> 00:03:20,420
I'm going to go ahead and
put in a little for loop,

70
00:03:20,420 --> 00:03:22,210
that's going to iterate

71
00:03:22,210 --> 00:03:26,070
through my new Counter object's items.

72
00:03:26,070 --> 00:03:28,100
So it is still a dictionary

73
00:03:28,100 --> 00:03:31,510
and therefore I can still
call the items method on it.

74
00:03:31,510 --> 00:03:33,370
And, the items method, of course,

75
00:03:33,370 --> 00:03:36,850
is going to give me back
key value pairs as tuples.

76
00:03:36,850 --> 00:03:39,600
I'm going to take those
tuples and pass them off

77
00:03:39,600 --> 00:03:42,180
to the sorted built-in function,

78
00:03:42,180 --> 00:03:46,860
which will give me those tuples
in lexicographical order.

79
00:03:46,860 --> 00:03:48,330
These are all lower-case strings

80
00:03:48,330 --> 00:03:50,950
so they will be in alphabetical order.

81
00:03:50,950 --> 00:03:53,760
And, for each word and
corresponding count,

82
00:03:53,760 --> 00:03:55,610
we're going to display a line of text

83
00:03:55,610 --> 00:03:57,790
containing the word and the count.

84
00:03:57,790 --> 00:03:59,820
So let's go ahead and execute that.

85
00:03:59,820 --> 00:04:02,520
There are all of the words
and all of the counts,

86
00:04:02,520 --> 00:04:04,328
and if you were to go
back and compare them

87
00:04:04,328 --> 00:04:07,200
to the output of the preceding script,

88
00:04:07,200 --> 00:04:08,660
where we used the same text,

89
00:04:08,660 --> 00:04:11,770
you'll see we got the same exact results.

90
00:04:11,770 --> 00:04:15,000
Now, separately we can also go ahead,

91
00:04:15,000 --> 00:04:16,870
like we did in the preceding script,

92
00:04:16,870 --> 00:04:21,010
and display the number of
unique keys once again.

93
00:04:21,010 --> 00:04:23,871
So to do that we're
going to use the length

94
00:04:23,871 --> 00:04:26,910
of the counter.keys method call.

95
00:04:26,910 --> 00:04:30,950
So counter.keys is going to
give us the actual collection

96
00:04:30,950 --> 00:04:34,220
of keys which is a set of unique values,

97
00:04:34,220 --> 00:04:35,600
and the length of that, of course,

98
00:04:35,600 --> 00:04:37,300
is the number of unique keys.

99
00:04:37,300 --> 00:04:39,090
We can also simply take the length

100
00:04:39,090 --> 00:04:40,780
of the Counter object itself

101
00:04:40,780 --> 00:04:42,630
because it is a dictionary

102
00:04:42,630 --> 00:04:44,290
and the number of key value pairs

103
00:04:44,290 --> 00:04:47,260
would also be the number of unique keys.

104
00:04:47,260 --> 00:04:49,000
So as you can see once again,

105
00:04:49,000 --> 00:04:51,213
the number of unique keys is 10.