1
00:00:00,090 --> 00:00:00,420
Okay.

2
00:00:00,420 --> 00:00:01,180
Welcome back.

3
00:00:01,200 --> 00:00:02,480
So we're finally here.

4
00:00:02,490 --> 00:00:04,500
This is a bit of a turning point in the course.

5
00:00:04,950 --> 00:00:12,180
Really important section not to scare you too much, but really fundamental to how SQL and my SQL are

6
00:00:12,180 --> 00:00:13,470
used in the real world.

7
00:00:14,200 --> 00:00:19,930
So this is a section all about relationships and also something called joints, which we'll get to in

8
00:00:19,930 --> 00:00:21,730
the second part of this section.

9
00:00:21,880 --> 00:00:24,460
So let's focus on the relationships part first.

10
00:00:24,610 --> 00:00:31,270
So far in this course, we've been working with very simple data, things like books or users where

11
00:00:31,270 --> 00:00:36,430
we only have a couple of columns per table and that's been by design.

12
00:00:36,430 --> 00:00:42,820
We need that simplicity in order to focus on what's important, to isolate the different features that

13
00:00:42,820 --> 00:00:48,130
we're talking about, whether it's functions or logical operators, it's much easier to teach those,

14
00:00:48,130 --> 00:00:54,070
and we don't have 20 different tables to worry about, and we can just basically drill down into one

15
00:00:54,070 --> 00:00:55,780
table and focus on that.

16
00:00:56,140 --> 00:01:02,890
But that's typically not how the real world works, where, yeah, we might have a user's table, but

17
00:01:02,890 --> 00:01:08,430
there's no site that really can just get by on having users unless it's featureless.

18
00:01:08,440 --> 00:01:14,590
There is plenty of other things that need to go along with users, whether it's comments or likes or

19
00:01:14,590 --> 00:01:19,300
tweets or posts or photos or any sort of, I don't know, order transaction.

20
00:01:19,300 --> 00:01:22,360
There are all these different things that we could have alongside users.

21
00:01:22,360 --> 00:01:27,340
So we've been working with simple data, but that's about to change.

22
00:01:29,330 --> 00:01:35,000
Real world data, unlike what we've been working with, is really messy, typically and interrelated.

23
00:01:35,000 --> 00:01:37,220
There's a lot of connections between data.

24
00:01:37,220 --> 00:01:42,440
So some of those things I talked about with, let's say, users, if we're working with just a simple

25
00:01:42,440 --> 00:01:47,840
blog site, so nothing like a Facebook or a complex social network, just a simple blog, we still need

26
00:01:47,840 --> 00:01:48,950
to have user data.

27
00:01:48,980 --> 00:01:52,610
We need to have information about posts, blog posts.

28
00:01:52,610 --> 00:01:56,630
We also need to somehow keep track of comments and then tags.

29
00:01:56,990 --> 00:01:59,120
And that can get pretty complicated already.

30
00:01:59,120 --> 00:02:05,060
Not to mention things like advertising and tracking not only ads on their own, but how ads are related

31
00:02:05,060 --> 00:02:07,940
to users and what users are clicking on and not clicking on.

32
00:02:07,940 --> 00:02:13,670
There's so much data that just from a simple blog site, something that may not seem that complex.

33
00:02:13,670 --> 00:02:16,130
There's so much stuff that we could store.

34
00:02:16,250 --> 00:02:22,430
So we're going to focus on how we work with interconnected data in the next couple of sections.

35
00:02:22,430 --> 00:02:24,620
So we're no longer focusing on one table.

36
00:02:24,650 --> 00:02:28,940
We're going to see how do we work with two tables or three or four that are related.

37
00:02:28,940 --> 00:02:34,880
So not just tables that exist independently, but tables that reference one another and are interconnected.

38
00:02:34,880 --> 00:02:35,960
So let's get going.

39
00:02:35,960 --> 00:02:39,830
And the first thing that I want to do is go back to our books data.

40
00:02:40,100 --> 00:02:42,110
So remember our books table.

41
00:02:42,140 --> 00:02:43,220
Don't worry, we're done with it.

42
00:02:43,220 --> 00:02:47,210
We're not going to be typing any books code, but on its own it was really simple.

43
00:02:47,210 --> 00:02:54,110
We had a books table with author, first name and last name, a book title released here, a page count

44
00:02:54,110 --> 00:02:58,790
in a stock inventory, I believe a stock quantity, and I think that was it.

45
00:02:59,150 --> 00:03:04,100
So again, on its own, that just represents a single book, but you can't really do much with that.

46
00:03:04,100 --> 00:03:09,440
But if we are running a sort of book site where we sold books, at a minimum, we would need to keep

47
00:03:09,440 --> 00:03:11,510
track of things like versions.

48
00:03:11,630 --> 00:03:13,130
So we didn't even talk about this.

49
00:03:13,130 --> 00:03:16,010
But there's often multiple versions of a single book.

50
00:03:16,010 --> 00:03:22,400
If you go on Amazon and you look at a book, let's say Harry Potter, the second book, Chamber of Secrets,

51
00:03:22,400 --> 00:03:27,650
well, there's going to be the initial British release, the UK version as a hardcover, and then there's

52
00:03:27,650 --> 00:03:31,220
a paperback, and then there's the American version, and then there's a paperback.

53
00:03:31,220 --> 00:03:36,050
There is a full color illustrated version that came out relatively, relatively recently.

54
00:03:36,050 --> 00:03:39,200
It looks pretty cool that has these giant illustrations.

55
00:03:39,200 --> 00:03:40,490
It's the same title.

56
00:03:41,240 --> 00:03:42,740
How do you keep track of that?

57
00:03:42,740 --> 00:03:46,940
Is that a different table or do you just make a separate book for every one of those?

58
00:03:46,940 --> 00:03:49,790
But if you do that, then how do you make sure that they're connected?

59
00:03:49,790 --> 00:03:54,890
Because when I go on Amazon and I look at the hardcover, it also will show me below.

60
00:03:55,160 --> 00:04:02,540
Some people also buy paperback version, or it's also available in, I don't know, 25, 40 different

61
00:04:02,540 --> 00:04:03,320
languages.

62
00:04:03,320 --> 00:04:06,710
So there's all sorts of things to worry about just with versions.

63
00:04:07,700 --> 00:04:09,050
But then we have authors.

64
00:04:09,350 --> 00:04:11,120
Well, that might seem straightforward.

65
00:04:11,120 --> 00:04:17,360
We have an author, first name and last name, two columns on our books table, but that's not sufficient

66
00:04:17,360 --> 00:04:18,980
when we have multiple authors.

67
00:04:19,519 --> 00:04:22,310
So some books, a lot of books have multiple authors.

68
00:04:22,310 --> 00:04:28,370
What about if it's a research paper that we're selling or a journal that has dozens of authors potentially?

69
00:04:28,910 --> 00:04:32,750
So there's a lot of things to keep track of there that we can't really do right now.

70
00:04:32,750 --> 00:04:35,510
We only have room for one author in our columns.

71
00:04:36,630 --> 00:04:39,270
Then a whole big thing is customers.

72
00:04:39,690 --> 00:04:44,910
If we're doing a book shop online, books on their own is great, but we also need ways to keep track

73
00:04:44,910 --> 00:04:48,060
of users or customers is what I'm calling them in this case.

74
00:04:48,240 --> 00:04:55,830
So customers on their own also can't do that much, but we might keep track of their email log in information.

75
00:04:56,280 --> 00:05:00,420
Of course, we wouldn't just store their password directly, but for simplicity's sake, let's say we're

76
00:05:00,420 --> 00:05:03,780
keeping track of password and email so that they can log in.

77
00:05:04,140 --> 00:05:06,460
And then there's a whole bunch of other things.

78
00:05:06,480 --> 00:05:07,230
Orders.

79
00:05:07,230 --> 00:05:11,050
So if customers want to order books, how do we store that?

80
00:05:11,130 --> 00:05:16,530
We need to create a new order and it needs to be associated with a customer who created it.

81
00:05:16,920 --> 00:05:18,750
It needs to have address information.

82
00:05:18,750 --> 00:05:21,630
Maybe it needs to have the contents of the order.

83
00:05:21,870 --> 00:05:22,920
Are there books?

84
00:05:22,920 --> 00:05:23,920
One book?

85
00:05:23,940 --> 00:05:24,830
How do we support that?

86
00:05:24,840 --> 00:05:25,890
How do we handle that?

87
00:05:26,340 --> 00:05:33,480
What about if we're working with dates and times and we need to keep track of when a return period expires?

88
00:05:33,480 --> 00:05:39,840
So we need to have a an order date and then keep track of 30 days or 60 days after that or whatever

89
00:05:39,840 --> 00:05:40,380
it is.

90
00:05:40,500 --> 00:05:47,550
Then we also need to store or somehow keep track of the transaction information addresses for billing

91
00:05:47,550 --> 00:05:51,520
address, not only shipping address, the type of shipping, the billing information.

92
00:05:51,540 --> 00:05:53,040
There's so much to store.

93
00:05:53,910 --> 00:05:56,270
And then a whole other thing is reviews.

94
00:05:56,280 --> 00:06:02,700
So if we go on Amazon or pretty much any book site, any book that you look at will have dozens or hundreds,

95
00:06:02,700 --> 00:06:04,410
sometimes thousands of reviews.

96
00:06:04,650 --> 00:06:09,090
So we've got reviews, their ratings, that's the number of stars.

97
00:06:09,090 --> 00:06:15,840
But then also a title for the review and then a caption and then the actual content of the review,

98
00:06:15,840 --> 00:06:18,060
which may be ten characters.

99
00:06:18,240 --> 00:06:20,850
I hate this book, however many characters that is.

100
00:06:20,850 --> 00:06:24,950
Or it could be a giant paragraph, someone's love letter that they're writing to a book.

101
00:06:24,960 --> 00:06:26,730
So how do we account for all of that?

102
00:06:26,730 --> 00:06:29,700
And then the last thing that we'll talk about are genres.

103
00:06:30,060 --> 00:06:34,990
So you may think that we could just get away with storing this in a book, which we absolutely could.

104
00:06:35,010 --> 00:06:36,570
We could just have one genre.

105
00:06:36,570 --> 00:06:39,900
But many books don't cleanly fit into one genre.

106
00:06:40,230 --> 00:06:46,170
We might have a vampire novel that's also historical fiction because it's set in Victorian period.

107
00:06:46,170 --> 00:06:54,300
England Or we might have a sci fi book that has elements of fantasy or who knows, you can have all

108
00:06:54,300 --> 00:06:59,460
sorts of cross-pollinated genres, so only assigning one genre might not be that useful.

109
00:06:59,460 --> 00:07:01,110
We may want to have a bunch more.

110
00:07:01,110 --> 00:07:05,940
And in fact, on Amazon you can see there are books that will have two or three different genres or

111
00:07:05,940 --> 00:07:07,830
different tags associated with them.

112
00:07:08,070 --> 00:07:12,600
So what I'm trying to show you here is that we were working with books on their own, but in a typical

113
00:07:12,600 --> 00:07:18,930
app or a typical website that has to do with books and a bookstore, in our case, there's a lot more

114
00:07:19,080 --> 00:07:23,280
that goes along with books to make anything functional, to store any important information.

115
00:07:23,280 --> 00:07:28,500
And this is, I don't know, maybe half of the tables that you would need to get away with kind of the

116
00:07:28,560 --> 00:07:30,240
the MVP or the bare minimum.

117
00:07:30,240 --> 00:07:32,670
There's still many other things.

118
00:07:32,670 --> 00:07:36,870
Something like orders, for instance, might actually need to be split into a couple of tables.

119
00:07:36,870 --> 00:07:41,070
We might need to have an address section or an address table.

120
00:07:41,070 --> 00:07:44,460
We might need to then have a separate billing table.

121
00:07:44,520 --> 00:07:46,950
We might need to have shipping.

122
00:07:46,950 --> 00:07:49,560
There could be all different tables that we could split it up into.

123
00:07:49,590 --> 00:07:52,920
So this isn't to say this is the end all, be all this is just that.

124
00:07:52,920 --> 00:07:56,730
There's a couple of other tables that we need to consider at the bare minimum.

125
00:07:56,880 --> 00:07:57,290
Okay.

126
00:07:57,330 --> 00:07:59,550
So we'll stop here in the next video.

127
00:07:59,550 --> 00:08:04,620
We're going to identify some of the different types of relationships between data at a high level,

128
00:08:04,620 --> 00:08:07,080
what are the different types of associations?

129
00:08:07,440 --> 00:08:08,520
And I'm back.

130
00:08:08,700 --> 00:08:10,500
I totally forgot to show you this.

131
00:08:10,500 --> 00:08:13,770
Even though I wrote a note to myself, I totally skipped this.

132
00:08:13,770 --> 00:08:17,130
So I'm back after recording this, adding this to the end.

133
00:08:17,880 --> 00:08:23,430
I want to show the schema diagram that I really think highlights how messy data can be.

134
00:08:23,850 --> 00:08:27,000
So here's I think I showed this early on in the course.

135
00:08:27,090 --> 00:08:31,770
This is just a potential schema for Facebook and it's definitely incomplete.

136
00:08:31,770 --> 00:08:35,039
It's maybe a 10th of the number of tables that Facebook has.

137
00:08:35,039 --> 00:08:42,600
But you can see just to store basic things like events and users, we have an event, Facebook event,

138
00:08:42,600 --> 00:08:49,800
we have event membership which is related to RSVP states, which is also related to venues, which is

139
00:08:49,800 --> 00:08:53,670
also related to groups and to workplaces and to Facebook profiles.

140
00:08:53,670 --> 00:08:55,500
And it just goes on and on and on.

141
00:08:55,890 --> 00:08:58,710
So what I'm trying to show you is that data can get messy.

142
00:08:58,710 --> 00:09:02,640
And in fact, here's another super intimidating schema.

143
00:09:02,670 --> 00:09:08,460
This one is a little bit crazy because it has to do with storing protein database.

144
00:09:08,610 --> 00:09:14,590
I think it was I saw some article on IBM and I believe MIT, and fortunately it's very low quality image.

145
00:09:14,590 --> 00:09:16,080
So you can actually see the tables.

146
00:09:16,080 --> 00:09:22,950
But each one of these little rectangles is a separate table and every line is a connection between the

147
00:09:22,950 --> 00:09:23,520
tables.

148
00:09:23,790 --> 00:09:25,980
Now, this is the worst I've ever seen, the craziest.

149
00:09:25,980 --> 00:09:30,300
And I think that was the point, is that they're trying to just show you they're never going to actually

150
00:09:30,300 --> 00:09:32,670
look at this to try and understand anything.

151
00:09:32,670 --> 00:09:36,210
But they're trying to show you how complex protein information is when you have.

152
00:09:36,320 --> 00:09:39,560
All these different combinations and permutations for a given protein.

153
00:09:39,680 --> 00:09:44,060
Without going into too much detail here, this isn't a bio, of course, but the way that a protein

154
00:09:44,060 --> 00:09:49,910
is folded, the same structure, basically the same amino acids, depending on how it's folded, can

155
00:09:49,910 --> 00:09:53,390
drastically change the features, if you will.

156
00:09:53,420 --> 00:09:55,820
Basically, what the protein does, how it behaves.

157
00:09:55,820 --> 00:10:05,030
And so this is a way of documenting using SQL in some way the structure of a protein using a database.

158
00:10:05,570 --> 00:10:07,100
So not to scare you there.

159
00:10:07,100 --> 00:10:08,900
I just think it's kind of a powerful image.

160
00:10:09,350 --> 00:10:17,510
Although grainy, it is powerful to see basically some really messy, ugly gross data that can be supported

161
00:10:17,750 --> 00:10:18,590
using tables.

162
00:10:18,590 --> 00:10:20,270
It's just doesn't look pretty.

163
00:10:20,420 --> 00:10:21,380
But that's the important part.

164
00:10:21,380 --> 00:10:26,240
And kind of the point of this video is I wanted to get you prepared mentally for working with some more

165
00:10:26,240 --> 00:10:28,130
complex data, although nothing that looks like this.

