1 00:00:05,280 --> 00:00:10,640 In this video, we'll have a look at the last of the four set operations, symmetric difference. 2 00:00:10,640 --> 00:00:15,600 The example is going to be a bit contrived, because I'm trying to make it non-maths related. 3 00:00:15,600 --> 00:00:20,720 Outside of mathematics, you may not need to find the symmetric difference of sets very often. 4 00:00:20,720 --> 00:00:27,455 Python supports the operation, so we'll have a look at what it can do, and it's there if you need it. 5 00:00:27,455 --> 00:00:30,321 For this example, we've got two sets. 6 00:00:30,321 --> 00:00:33,680 They represent programming courses at a training establishment. 7 00:00:33,680 --> 00:00:38,280 We'd like to learn more than one language, and we're determined not to miss any lessons. 8 00:00:39,440 --> 00:00:44,000 That means we'll have to take one course in the mornings, and another in the afternoons. 9 00:00:44,000 --> 00:00:48,707 Which means we can't take any course that has lessons in both the morning and the afternoon. 10 00:00:51,040 --> 00:00:54,880 We've seen the set intersection before. We can use that to see which courses 11 00:00:54,880 --> 00:00:59,120 have lessons in the morning and afternoon. The green area represents the intersection 12 00:00:59,120 --> 00:01:02,934 of the two sets, and shows us which courses we can't take. 13 00:01:04,319 --> 00:01:09,360 There are only 6 courses in this example, so it's not hard to see which courses we can take. 14 00:01:09,440 --> 00:01:15,280 Bear with me on this, and imagine sets that could contain hundreds (or thousands) of items. 15 00:01:15,280 --> 00:01:19,520 To get the set of courses we can take, we want the opposite of the intersection. 16 00:01:22,240 --> 00:01:25,440 The symmetric difference is the opposite of intersection. 17 00:01:25,440 --> 00:01:30,400 It produces the set of items that are in one set, or the other, but not in both. 18 00:01:30,400 --> 00:01:36,800 Java, C, C#, and Ruby are in both sets, so they're excluded from the symmetric difference. 19 00:01:36,800 --> 00:01:40,480 That leaves Python and Lisp, which is exactly what we wanted. 20 00:01:40,480 --> 00:01:44,201 So lets see how we produce the symmetric difference in code. 21 00:01:45,040 --> 00:01:49,680 I've created a new Python file called set_sd.py for this example. 22 00:01:49,680 --> 00:01:52,867 The name set_symmetric_difference is a bit long, 23 00:01:52,867 --> 00:01:58,561 so I've abbreviated it. Pause the video, while you create your Python file. 24 00:01:59,120 --> 00:02:02,094 We start with the two sets that contain the courses: 25 00:02:39,040 --> 00:02:43,880 The symmetric difference operator is the ^ symbol, and I'll start with the operator: 26 00:02:50,640 --> 00:02:54,080 We could also use the method, and we'll do that next. 27 00:02:54,080 --> 00:02:57,254 First, I'll print the possible courses that we can take: 28 00:03:03,897 --> 00:03:05,454 and run the program. 29 00:03:08,480 --> 00:03:12,560 There shouldn't be any surprises with the output, we get Lisp and Python. 30 00:03:12,560 --> 00:03:17,600 Trying to do that, by iterating over the sets, would be quite fiddly. Have a go, 31 00:03:17,600 --> 00:03:22,480 if you want to check how messy the code would be. I won't type the code to do that, on video, 32 00:03:22,480 --> 00:03:28,000 because you just wouldn't do it that way. At least, I don't recommend that you do it that way. 33 00:03:28,000 --> 00:03:31,360 Even if these courses were stored in lists, rather than sets, 34 00:03:31,360 --> 00:03:36,400 you can still produce the symmetric difference. You'd convert one list to a set, and pass the 35 00:03:36,400 --> 00:03:41,894 other list to the symmetric_difference method. That's worth seeing, so I will do that. 36 00:03:42,640 --> 00:03:47,574 First, we change lines 1 and 2 so that we've got lists, rather than sets. 37 00:03:56,000 --> 00:04:01,187 We've now got a warning from IntelliJ, because the ^ operator isn't defined for lists. 38 00:04:01,187 --> 00:04:04,142 We have to use the method instead: 39 00:04:10,265 --> 00:04:15,401 That's removed the warning. The symmetric difference operator is defined for the set. 40 00:04:15,401 --> 00:04:17,440 But we've got another warning now. 41 00:04:17,440 --> 00:04:22,960 We can't use the operator with a set and a list. We could convert afternoon to a set as well. 42 00:04:22,960 --> 00:04:27,840 Or we could use the method instead. I'll change the code to use the method: 43 00:04:39,520 --> 00:04:44,720 As we've seen, an advantage of the methods is that we can pass any iterable to them. That avoids the 44 00:04:44,720 --> 00:04:50,054 need to explicitly convert afternoon to a set. Ok, run the program: 45 00:04:51,920 --> 00:04:55,920 and we get the same result. Remember that the order might be different, 46 00:04:55,920 --> 00:05:00,000 because sets have no ordering. Unlike difference, symmetric 47 00:05:00,000 --> 00:05:04,080 difference is commutative – you can perform the operation either way round. 48 00:05:04,080 --> 00:05:06,790 So we could also write that as: 49 00:05:27,481 --> 00:05:30,201 Run the program again 50 00:05:30,800 --> 00:05:35,440 and once again, we get the same result. As with the other set operations, 51 00:05:35,440 --> 00:05:41,840 there's also a symmetric_difference_update operation. The operator is ^=, if you prefer to 52 00:05:41,840 --> 00:05:46,640 use the operators rather than call the methods. And that's symmetric difference. 53 00:05:46,640 --> 00:05:49,053 If you find that you need to do something like this, 54 00:05:49,053 --> 00:05:54,000 the symmetric difference operation will save a lot of code – and reduce the risk of adding a bug. 55 00:05:54,080 --> 00:05:58,560 It's one more tool in your toolkit. And if you do find a non-mathematical use for it, 56 00:05:58,560 --> 00:06:02,360 please let us know in the Q&A. In the next few videos, 57 00:06:02,360 --> 00:06:06,640 we'll look at super-sets and sub-sets. See you in the next video.