1
00:00:00,000 --> 00:00:03,012
Throughout my experience with working with large data

2
00:00:03,013 --> 00:00:06,762
sets, I've noticed some techniques that data collecting

3
00:00:06,763 --> 00:00:10,836
institutes use to save storage and make them

4
00:00:10,837 --> 00:00:12,804
their data have less space

5
00:00:12,805 --> 00:00:15,338
on computer disks. A common technique

6
00:00:15,339 --> 00:00:17,226
is not to use decimals.

7
00:00:17,227 --> 00:00:19,066
For example, if you search for temperature

8
00:00:19,067 --> 00:00:21,674
data, you'll not see the actual temperatures.

9
00:00:21,675 --> 00:00:23,034
Let's say these are celsius.

10
00:00:23,035 --> 00:00:25,164
You'll not see them like that, but that

11
00:00:25,165 --> 00:00:27,292
you'll see them without the dots because that

12
00:00:27,293 --> 00:00:29,890
will save storage in the servers.

13
00:00:29,891 --> 00:00:33,628
So one character less, that means some bytes less.

14
00:00:33,629 --> 00:00:38,140
So your duty as a programmer is to actually divide all

15
00:00:38,141 --> 00:00:42,112
those values by 10, so that you get the actual data.

16
00:00:42,113 --> 00:00:45,648
So let's say these data were in some text files, and

17
00:00:45,649 --> 00:00:49,222
we open these data with Python, and the course actually covers

18
00:00:49,223 --> 00:00:52,486
how to manage, how to open text files in Python.

19
00:00:52,487 --> 00:00:54,580
But let us skip that part for now.

20
00:00:54,581 --> 00:00:57,460
So let's say we got them loaded as a list here.

21
00:00:57,461 --> 00:01:01,076
Now we want to actually divide by 10 any of them.

22
00:01:01,077 --> 00:01:02,950
So how do you do that?

23
00:01:02,951 --> 00:01:07,278
Well, a technique is to iterate over temps.

24
00:01:07,279 --> 00:01:12,950
So for temp in temps, and in each iteration,

25
00:01:12,951 --> 00:01:16,668
you want to do temp divided by ten.

26
00:01:16,669 --> 00:01:20,908
But where do we store this temp divided by ten?

27
00:01:20,909 --> 00:01:23,530
Well, we can store it in a new list.

28
00:01:23,531 --> 00:01:25,110
Let's say new_temps.

29
00:01:26,010 --> 00:01:28,080
That is an empty list for now.

30
00:01:28,081 --> 00:01:31,515
So the loop will iterate through this list, and in each iteration,

31
00:01:31,516 --> 00:01:33,700
[Author typing]

32
00:01:33,701 --> 00:01:40,530
we going to append this value to new_temps.

33
00:01:40,531 --> 00:01:46,452
So append this value. That looks okay,

34
00:01:46,453 --> 00:01:49,970
and then print(new temps).

35
00:01:49,971 --> 00:01:52,240
Let me go ahead and execute that.

36
00:01:52,241 --> 00:01:54,566
[Author typing]

37
00:01:54,566 --> 00:01:57,086
So this is the desired output.

38
00:01:57,087 --> 00:01:59,368
Now, that is very correct.

39
00:01:59,369 --> 00:02:02,152
But there is a neater way to do

40
00:02:02,153 --> 00:02:05,064
this in just one line of Python code,

41
00:02:05,065 --> 00:02:07,954
and that is by using list comprehensions.

42
00:02:07,955 --> 00:02:09,639
It goes like this.

43
00:02:10,810 --> 00:02:13,004
You don't have to create an empty list

44
00:02:13,005 --> 00:02:16,130
because a list will be generated dynamically.

45
00:02:16,131 --> 00:02:23,334
You'd say temp / 10 for temp in temps,

46
00:02:24,833 --> 00:02:28,624
print(new_temps) to see what we get,

47
00:02:28,625 --> 00:02:31,950
execute, and you get exactly the same output.

48
00:02:33,170 --> 00:02:35,898
So that is a list comprehension.

49
00:02:35,899 --> 00:02:39,332
It's a way to build lists without having to

50
00:02:39,333 --> 00:02:42,932
create a for loop, a standard for loop, because

51
00:02:42,933 --> 00:02:46,440
we actually have an inline for loop in here.

52
00:02:46,441 --> 00:02:49,208
So what's going on here is that we have an

53
00:02:49,209 --> 00:02:52,520
iteration here, and in each iteration we're going to store

54
00:02:52,521 --> 00:02:56,428
this value in the list, in the new_temps list.

55
00:02:56,429 --> 00:03:01,292
So this is like saying store temp divided by ten,

56
00:03:01,293 --> 00:03:04,594
but Python will say, okay, but what is temp?

57
00:03:04,595 --> 00:03:06,490
It's a new variable.

58
00:03:06,491 --> 00:03:08,258
We haven't defined it anywhere.

59
00:03:08,259 --> 00:03:11,794
And then we say, well, for temp in temps.

60
00:03:11,795 --> 00:03:13,970
So temp is a variable of temps.

61
00:03:13,971 --> 00:03:18,076
So that's how you create a list with these

62
00:03:18,077 --> 00:03:21,470
values for each of the values of temps.

63
00:03:22,370 --> 00:03:24,270
That is a list comprehension.

64
00:03:24,271 --> 00:03:27,766
[Outro sound]