1
00:00:00,005 --> 00:00:04,000
- [Instructor] Reading a CSV file as an array of arrays

2
00:00:04,000 --> 00:00:05,004
isn't the only option.

3
00:00:05,004 --> 00:00:08,007
You can also read the data into a dictionary structure,

4
00:00:08,007 --> 00:00:12,004
and you can do this by using the DictReader class,

5
00:00:12,004 --> 00:00:14,007
which is also in the CSV module,

6
00:00:14,007 --> 00:00:16,000
so I will encourage you

7
00:00:16,000 --> 00:00:18,007
to refer to the documentation link for that.

8
00:00:18,007 --> 00:00:23,006
So let's open up our sample file, read_csv_dict.py,

9
00:00:23,006 --> 00:00:26,009
and what we're going to do is modify our original example

10
00:00:26,009 --> 00:00:30,007
to use the dictionary reader instead of the default setting.

11
00:00:30,007 --> 00:00:34,006
So my starting point code defines an empty dictionary,

12
00:00:34,006 --> 00:00:38,007
rather than an empty array that's going to hold the data.

13
00:00:38,007 --> 00:00:42,007
So once again, we're going to open up the file in read mode.

14
00:00:42,007 --> 00:00:44,006
So this should look familiar by now,

15
00:00:44,006 --> 00:00:51,007
so I'll open up the file name in read mode as CSV file,

16
00:00:51,007 --> 00:00:52,005
but this time,

17
00:00:52,005 --> 00:00:56,000
we're going to be using the DictReader class.

18
00:00:56,000 --> 00:01:04,008
So I'll type CSV DictReader and we'll pass in the CSV file.

19
00:01:04,008 --> 00:01:08,003
Alright, and then we're going to loop over each row

20
00:01:08,003 --> 00:01:09,007
and read one at a time,

21
00:01:09,007 --> 00:01:13,002
just as we did for the array example.

22
00:01:13,002 --> 00:01:19,007
So for row in reader.

23
00:01:19,007 --> 00:01:20,006
Now in this case,

24
00:01:20,006 --> 00:01:23,001
we're going to add the row to our dictionary,

25
00:01:23,001 --> 00:01:27,005
which requires a key for the row to go into.

26
00:01:27,005 --> 00:01:31,003
So I'm going to use the first value in the row as the key,

27
00:01:31,003 --> 00:01:33,002
but how do I get that value?

28
00:01:33,002 --> 00:01:37,005
Because remember, the row at this point is not an array,

29
00:01:37,005 --> 00:01:41,003
it's a dictionary of key-value pairs.

30
00:01:41,003 --> 00:01:45,000
So the DictReader class has a property called field names,

31
00:01:45,000 --> 00:01:48,007
which it uses to map the values in each row,

32
00:01:48,007 --> 00:01:50,008
and this property is a list

33
00:01:50,008 --> 00:01:55,001
that contains the values in the CSV header row.

34
00:01:55,001 --> 00:02:02,002
So field names sub zero will be given the value item name

35
00:02:02,002 --> 00:02:03,009
if we look at the inventory data, right?

36
00:02:03,009 --> 00:02:08,002
So this is going to be 0, 1, 2, and so on,

37
00:02:08,002 --> 00:02:10,002
and we'll use that as the value,

38
00:02:10,002 --> 00:02:13,003
is the key into the row data to get the next item.

39
00:02:13,003 --> 00:02:15,003
It's a lot easier to see this in action,

40
00:02:15,003 --> 00:02:17,000
so let's just run an experiment.

41
00:02:17,000 --> 00:02:19,004
So before we do our loop,

42
00:02:19,004 --> 00:02:22,003
let's just print the first row in the data.

43
00:02:22,003 --> 00:02:26,008
So I'm going to write row equals next on the reader,

44
00:02:26,008 --> 00:02:30,007
and then print the row,

45
00:02:30,007 --> 00:02:35,005
and I'm going to also print the field names property.

46
00:02:35,005 --> 00:02:38,009
So I'll call reader.fieldnames, right?

47
00:02:38,009 --> 00:02:41,008
So before we actually do our loop,

48
00:02:41,008 --> 00:02:46,000
let's just go ahead and run what we have.

49
00:02:46,000 --> 00:02:50,002
So let's run this in the terminal.

50
00:02:50,002 --> 00:02:52,005
Right, okay, so when we run this,

51
00:02:52,005 --> 00:02:55,006
you can see that the first row is a dictionary

52
00:02:55,006 --> 00:02:58,004
consisting of key-value pairs,

53
00:02:58,004 --> 00:03:01,005
and the names property contains the values

54
00:03:01,005 --> 00:03:05,000
that are used as the keys in each row.

55
00:03:05,000 --> 00:03:07,003
So if I want to use the item name

56
00:03:07,003 --> 00:03:09,003
as the overall dictionary key,

57
00:03:09,003 --> 00:03:11,002
I just need to get the name of that key

58
00:03:11,002 --> 00:03:13,003
from the field names property.

59
00:03:13,003 --> 00:03:19,004
So now let's go back and write our loop.

60
00:03:19,004 --> 00:03:24,004
So for each row in the reader on the data,

61
00:03:24,004 --> 00:03:29,001
I'm going to get in the row,

62
00:03:29,001 --> 00:03:31,009
I'm going to get the reader field names,

63
00:03:31,009 --> 00:03:34,003
and I'm going to use the product name.

64
00:03:34,003 --> 00:03:41,009
That's going to be at index zero is equal to the row of data.

65
00:03:41,009 --> 00:03:43,007
Alright, so now this sets that row of data

66
00:03:43,007 --> 00:03:45,009
into my dictionary.

67
00:03:45,009 --> 00:03:51,001
I no longer need the print statements,

68
00:03:51,001 --> 00:03:53,001
and I'm going to leave this next in place

69
00:03:53,001 --> 00:03:55,002
'cause I want to skip over the headers.

70
00:03:55,002 --> 00:03:57,000
I don't want the headers to be part of the data.

71
00:03:57,000 --> 00:04:01,005
So I want to skip that row and just do the actual data rows.

72
00:04:01,005 --> 00:04:04,009
I've already got my return data in place,

73
00:04:04,009 --> 00:04:06,003
and as we did earlier,

74
00:04:06,003 --> 00:04:08,005
let's go ahead and print out some results.

75
00:04:08,005 --> 00:04:11,004
I'm going to use the Python Pretty Print module,

76
00:04:11,004 --> 00:04:14,003
and rather, instead of the regular print statement

77
00:04:14,003 --> 00:04:15,007
for this example,

78
00:04:15,007 --> 00:04:20,000
so I'm going to write pprint.pprint,

79
00:04:20,000 --> 00:04:24,002
and I'm going to print out the inventory data.

80
00:04:24,002 --> 00:04:26,008
And then let's also do the same thing,

81
00:04:26,008 --> 00:04:32,002
I'm going to print out the inventory data for the key of Apple,

82
00:04:32,002 --> 00:04:33,006
and then let's also print out

83
00:04:33,006 --> 00:04:39,007
the consumer price for the apple.

84
00:04:39,007 --> 00:04:41,007
Alright.

85
00:04:41,007 --> 00:04:46,005
Okay, let's save that and then let's run the code.

86
00:04:46,005 --> 00:04:49,000
I'm going to run this in the terminal,

87
00:04:49,000 --> 00:04:51,002
and let's make the terminal bigger.

88
00:04:51,002 --> 00:04:53,009
Alright, so you can see that when I run the code,

89
00:04:53,009 --> 00:04:57,003
the data is now in a dictionary, right?

90
00:04:57,003 --> 00:05:02,005
And the key for each one of the values is the item name.

91
00:05:02,005 --> 00:05:04,006
So starting all the way up here,

92
00:05:04,006 --> 00:05:08,000
we've got avocado, bagel, baguette, banana, so on.

93
00:05:08,000 --> 00:05:10,002
It looks like we're missing apple.

94
00:05:10,002 --> 00:05:11,000
You know what?

95
00:05:11,000 --> 00:05:13,005
Maybe it was a bad idea to leave that next statement in.

96
00:05:13,005 --> 00:05:16,008
Let's go ahead and just take that out,

97
00:05:16,008 --> 00:05:21,003
and let's run this again.

98
00:05:21,003 --> 00:05:25,001
Alright, let's scroll all the way up.

99
00:05:25,001 --> 00:05:25,009
Okay, there we go.

100
00:05:25,009 --> 00:05:27,007
Apple's there now, avocado,

101
00:05:27,007 --> 00:05:29,000
and you can see that each key

102
00:05:29,000 --> 00:05:32,004
has this data object associated with it,

103
00:05:32,004 --> 00:05:37,000
and let's scroll all the way back down to the bottom.

104
00:05:37,000 --> 00:05:38,006
Right, so here at the bottom

105
00:05:38,006 --> 00:05:42,003
where we accessed the specific item name

106
00:05:42,003 --> 00:05:43,004
using the Apple key,

107
00:05:43,004 --> 00:05:46,006
we can see that it has that particular row.

108
00:05:46,006 --> 00:05:48,009
So it's got the category fruits, and so on,

109
00:05:48,009 --> 00:05:51,002
and then the last line where we get the consumer price,

110
00:05:51,002 --> 00:05:53,007
you can see that that's 0.75.

111
00:05:53,007 --> 00:05:56,008
Now if this seems a bit complicated, okay,

112
00:05:56,008 --> 00:05:58,003
you don't need to do it that way,

113
00:05:58,003 --> 00:06:02,005
because remember, you can use other values for the keys.

114
00:06:02,005 --> 00:06:04,002
So let's modify the code

115
00:06:04,002 --> 00:06:08,007
to just use an integer for each key.

116
00:06:08,007 --> 00:06:12,004
Alright, so rather than using the item name,

117
00:06:12,004 --> 00:06:14,006
I'm going to make a variable called key.

118
00:06:14,006 --> 00:06:17,000
I'm going to set that to zero, alright?

119
00:06:17,000 --> 00:06:20,000
And then for each row in the reader,

120
00:06:20,000 --> 00:06:23,009
what I'm going to do is, instead of doing this,

121
00:06:23,009 --> 00:06:27,008
I'm just going to have the key, that index be the key,

122
00:06:27,008 --> 00:06:33,001
and then I'm going to say key +=1,

123
00:06:33,001 --> 00:06:35,009
and then obviously we need to change our print statements.

124
00:06:35,009 --> 00:06:39,000
So this is no longer getting the apple.

125
00:06:39,000 --> 00:06:44,006
This is getting, let use zero as the index, alright?

126
00:06:44,006 --> 00:06:50,004
So let's run again.

127
00:06:50,004 --> 00:06:53,008
Alright, and sure enough, now you can see

128
00:06:53,008 --> 00:06:56,006
that instead of the keys being the item names,

129
00:06:56,006 --> 00:06:58,002
I have these integer values,

130
00:06:58,002 --> 00:07:01,005
which are now serving as the keys, right?

131
00:07:01,005 --> 00:07:03,003
So we have the same result,

132
00:07:03,003 --> 00:07:06,006
only now the item name is not there anymore,

133
00:07:06,006 --> 00:07:08,007
but you can see that the output is essentially the same.

134
00:07:08,007 --> 00:07:10,002
So it's really up to you

135
00:07:10,002 --> 00:07:14,000
how you want to represent the keys in your data dictionary.