1
00:00:02,140 --> 00:00:07,770
So in the last lecture you saw how you can push elements into newly created arrays and I will emphasize

2
00:00:07,810 --> 00:00:12,930
this again, that the values you push don't have to be arrays themselves.

3
00:00:12,940 --> 00:00:16,990
We could have pushed the age or the name, that would have worked too

4
00:00:17,110 --> 00:00:22,360
and then we would have all hobbies or a different name then which would be an array of all the ages

5
00:00:22,360 --> 00:00:24,890
that are in the group or all the names that are in the group,

6
00:00:24,910 --> 00:00:27,030
so this would have worked too.

7
00:00:27,040 --> 00:00:33,440
I however want to have these values in all hobbies but not in their nested arrays

8
00:00:33,550 --> 00:00:40,600
and for that, I will add a new pipeline stage which we haven't seen before and that is the unwind stage.

9
00:00:41,020 --> 00:00:49,200
The unwind stage is always a great stage when you have an array of which you want to pull out the elements.

10
00:00:49,330 --> 00:00:56,500
Now to understand what unwind does, I'll temporarily move my grouping out of there so that I can just copy

11
00:00:56,500 --> 00:00:59,770
that later and I'll complete the unwind stage.

12
00:00:59,770 --> 00:01:05,500
Unwind has two different syntaxes and you can check the official docs for all the details as always,

13
00:01:06,280 --> 00:01:07,910
in its most common usage,

14
00:01:08,050 --> 00:01:11,800
you just pass the name of a field that holds an array,

15
00:01:12,010 --> 00:01:13,180
so hobbies,

16
00:01:13,180 --> 00:01:20,500
remember we get our own initial documents as an input to this stage and now, we just unwind on hobbies.

17
00:01:20,530 --> 00:01:22,780
Now what does unwind do?

18
00:01:23,140 --> 00:01:27,320
Well let's take that stage and let's see. If I execute this,

19
00:01:27,430 --> 00:01:33,370
you see what I get back are a couple of documents more than we originally had because now we have

20
00:01:33,370 --> 00:01:38,770
a document Max with hobbies sports and we have Max again with hobbies cooking.

21
00:01:38,770 --> 00:01:46,170
So what unwind does is it basically flattens the array by repeating the document

22
00:01:46,250 --> 00:01:54,730
that held the array as often as needed to merge it with the array elements. So the original array of Max 

23
00:01:54,970 --> 00:01:57,190
simply has sports and cooking,

24
00:01:57,190 --> 00:01:59,140
therefore Max was repeated twice,

25
00:01:59,170 --> 00:02:05,880
we get two new documents, Max with sports and Max with cooking and the same is true for all other elements.

26
00:02:05,950 --> 00:02:13,990
So where group merges multiple documents into one, unwind takes one document and spits out multiple

27
00:02:13,990 --> 00:02:14,930
documents

28
00:02:15,340 --> 00:02:17,850
and now with that, we can group again

29
00:02:17,890 --> 00:02:22,870
but now we can simply take hobbies which is no longer an array but a single value.

30
00:02:23,200 --> 00:02:27,540
So if I now bring back my group stage, I can leave it as it is

31
00:02:28,220 --> 00:02:32,570
but now you will see, if I repeat this,

32
00:02:32,710 --> 00:02:36,070
I still only have two groups because the ages haven't changed

33
00:02:36,310 --> 00:02:38,150
but now you see all hobbies holds

34
00:02:38,170 --> 00:02:41,710
just my array values and not the embedded arrays.

35
00:02:41,740 --> 00:02:46,240
However you can also see that we can have duplicate values,

36
00:02:46,240 --> 00:02:48,120
so what can we do about that?