1
00:00:00,300 --> 00:00:06,500
And we will come back to the request of our course about the complete introduction to The Science Show

2
00:00:06,540 --> 00:00:13,470
until now, we talked about statistics and I hope you guys have a good understanding of this topic.

3
00:00:14,250 --> 00:00:22,350
Now it's time to get a bit more into action and to learn about some tools that are used in data science

4
00:00:22,350 --> 00:00:29,070
and when we are analyzing data so it can also be used in machine learning as well as deep learning.

5
00:00:29,940 --> 00:00:32,550
So basically one of those tools is no.

6
00:00:32,970 --> 00:00:35,500
So in this class, we are going to have an introduction.

7
00:00:35,910 --> 00:00:39,300
We are not going to practice what it is going to be in the next few classes.

8
00:00:39,300 --> 00:00:45,210
But in this class, we are going to talk about what is no what's the difference between lumpy and lists

9
00:00:45,360 --> 00:00:52,240
and what are the advantages of using numpty in general versus Python lists, of course.

10
00:00:52,830 --> 00:00:53,650
So let's.

11
00:00:54,640 --> 00:00:58,820
So basically non P stands for numerical python.

12
00:00:59,250 --> 00:01:06,210
And what you guys need to understand is that it's a Python library that consists of multidimensional

13
00:01:06,240 --> 00:01:12,070
arrays, objects and a collection of routines for processing those areas.

14
00:01:12,450 --> 00:01:18,530
So basically numpty, as I said, it's it's a library in Python.

15
00:01:18,810 --> 00:01:23,910
So there are many features of them that, well, that he has.

16
00:01:24,720 --> 00:01:31,410
First of all, it will allow you multi will allow you using multidimensional arrays, which is pretty

17
00:01:31,410 --> 00:01:36,810
cool and will fester will will allow you to make faster operations.

18
00:01:36,810 --> 00:01:42,090
So if you need to make faster operations that will allow you to make faster operations, it has the

19
00:01:42,090 --> 00:01:44,390
functions and operators for those areas.

20
00:01:44,400 --> 00:01:49,890
So once again, you will be able to create one dimension, two dimensional, three dimensional areas

21
00:01:50,190 --> 00:01:56,130
and you will have different functions and operators that will be used with those areas too fast to make

22
00:01:56,130 --> 00:01:57,990
it faster for all the calculations.

23
00:01:58,980 --> 00:02:01,350
It's a good alternative for Matlab.

24
00:02:01,360 --> 00:02:10,950
So for the for those of you who know what that is, it's a python alternative for this, for Matlab

25
00:02:10,950 --> 00:02:11,520
in general.

26
00:02:13,570 --> 00:02:15,830
Also, it's a four year transform.

27
00:02:15,860 --> 00:02:22,560
It's well, the four year transforms and shape manipulation, so it will allow you to have shape manipulation.

28
00:02:22,570 --> 00:02:31,450
So when you create your 3D, multitenant, multidimensional areas, you will be able to well, to use

29
00:02:31,480 --> 00:02:34,690
the shapes and control the shape manipulation.

30
00:02:35,140 --> 00:02:40,780
And finally, there is linear algebra and random number generation that is integrated to it.

31
00:02:41,020 --> 00:02:49,000
And with linear algebra, you will be able to work faster your calculations because using multi multidimensional

32
00:02:49,000 --> 00:02:56,920
areas will allow you using linear algebra as well and faster this way your calculations.

33
00:02:58,810 --> 00:03:02,080
So once again, there are plenty of advantages of using them.

34
00:03:02,260 --> 00:03:06,710
So as I said, the first of all would be the area oriented computing.

35
00:03:07,030 --> 00:03:09,620
So those are areas that we are talking about right now.

36
00:03:09,640 --> 00:03:15,460
So this would be one dimensional area, two dimensional areas, and those are three dimensional areas.

37
00:03:15,460 --> 00:03:19,880
So here it's presented very, very simply.

38
00:03:20,140 --> 00:03:22,350
Usually the areas that are way bigger.

39
00:03:22,360 --> 00:03:29,440
So the amount of calculations that is here is well, that you will be able to do by using numpty is

40
00:03:29,440 --> 00:03:33,350
pretty huge versus working with the Python list.

41
00:03:33,370 --> 00:03:39,730
For example, it's the efficient, multidimensional, well, efficient, multidimensional area implementation.

42
00:03:39,760 --> 00:03:42,370
So this is exactly what I was talking about.

43
00:03:42,380 --> 00:03:45,120
So there are it's really, really efficient.

44
00:03:45,160 --> 00:03:48,400
There are multiple multi arrays.

45
00:03:48,670 --> 00:03:50,980
Multidimensional areas are pretty efficient.

46
00:03:50,980 --> 00:03:57,870
So it will allow you, as I said, to make huge calculations in a fast amount of time.

47
00:03:59,380 --> 00:04:02,330
It's really created for scientific computation also.

48
00:04:03,100 --> 00:04:06,370
So as I said, once again, it's four huge calculations.

49
00:04:06,940 --> 00:04:09,250
So it's not like Python list.

50
00:04:09,250 --> 00:04:10,950
You can use Python this once again.

51
00:04:10,960 --> 00:04:18,820
It will not be as efficient as using no faster resolution of operations on entire arrays of data without

52
00:04:18,820 --> 00:04:23,640
writing loops with the use of standardized mathematical functions.

53
00:04:23,890 --> 00:04:30,400
So as we said before, there is a linear algebra and some other functions and operations that are integrated

54
00:04:30,400 --> 00:04:37,270
with the number that will allow you to make huge operations on arrays well, on three dimensional areas.

55
00:04:37,270 --> 00:04:38,720
That will also be really huge.

56
00:04:39,250 --> 00:04:44,740
And this will save you a lot of times when you guys are making huge operations once again, maybe right

57
00:04:44,740 --> 00:04:48,610
now you don't see why you use this once again.

58
00:04:48,610 --> 00:04:55,780
If you guys create huge programs, the use of is is really recommended.

59
00:04:57,580 --> 00:05:00,790
There are different tools that are integrated with no, first of all.

60
00:05:00,970 --> 00:05:02,650
Well, you can read I read that.

61
00:05:02,890 --> 00:05:04,930
You can read, I read that on disk.

62
00:05:05,210 --> 00:05:07,120
You can write, I read that on disk.

63
00:05:07,390 --> 00:05:13,270
And you can work with different memory map files, which is a pretty good, pretty cool as well.

64
00:05:14,200 --> 00:05:17,600
As we said before, there is an integration of fully transform capabilities.

65
00:05:17,890 --> 00:05:24,850
And finally, as we said in the past, our point there is linear algebra and random number generation.

66
00:05:26,310 --> 00:05:29,710
All right, so finally, numpty versus versus Biton list.

67
00:05:29,740 --> 00:05:37,050
So, as I said, the number is a Python library and the three of the three main reasons of why using

68
00:05:37,050 --> 00:05:43,380
nonpeak, the first one will be it will consume less memory since there is less operations to do.

69
00:05:44,070 --> 00:05:48,090
It will go way faster and it will consume way, way less.

70
00:05:48,240 --> 00:05:56,140
Memory has an execution speed that well, this will bring us to the second point that is having an issue

71
00:05:56,230 --> 00:06:00,170
in the execution speed that is way faster than Python list.

72
00:06:00,540 --> 00:06:07,780
So I said it goes way faster than working with Python Lisk, especially if you guys want to work with

73
00:06:07,800 --> 00:06:08,910
the huge operations.

74
00:06:09,750 --> 00:06:12,390
Once again, this is an interesting thing that I found.

75
00:06:12,390 --> 00:06:16,140
So you can compare the speed of Arizona versus lists.

76
00:06:16,470 --> 00:06:20,550
So as you can see, well, the time is in milliseconds, so it's not even one second.

77
00:06:20,850 --> 00:06:29,220
But you still can see that number is at least four to five times faster than the list than Python list.

78
00:06:30,210 --> 00:06:35,120
So once again, there is a huge well, this is a huge difference, especially if you guys work with

79
00:06:35,140 --> 00:06:40,710
huge operations, if you can with numpty, for example, the result of an operation in one second,

80
00:06:40,920 --> 00:06:45,090
it will take you with Python list, maybe like five to 10 seconds to resolve.

81
00:06:45,120 --> 00:06:52,790
So if you have a set of operations, Numpty is really efficient way to be able to resolve all the operations.

82
00:06:54,030 --> 00:06:56,290
And finally, it's just convenient to use.

83
00:06:56,310 --> 00:07:02,070
So it's pretty much it's not well, you have to get used to it to use it properly.

84
00:07:02,460 --> 00:07:08,450
But it's not that complicated to use since it's a Python library, it's already integrated into Python.

85
00:07:09,540 --> 00:07:16,110
So you guys know exactly what what numpty number is, how?

86
00:07:16,110 --> 00:07:22,590
Well, maybe not how it works, but you know exactly how what what it is, why exactly we use it and

87
00:07:22,860 --> 00:07:27,070
why we prefer using numbers instead of python lists.

88
00:07:27,070 --> 00:07:32,080
So all the advantages and the features of this Python library.

89
00:07:32,700 --> 00:07:35,520
So in the next few classes, we are going to practice a little bit.

90
00:07:35,520 --> 00:07:38,590
Writing Numpty functions with Python, of course.

91
00:07:39,090 --> 00:07:46,190
So you will see what it looks like and you will have a certain introduction to this Python library.

92
00:07:46,500 --> 00:07:49,770
So that's it for this class, guys, and see all in our next class.