1 00:00:04,818 --> 00:00:10,543 There are three methods we can use, to delete items from a set. There's also a clear method, 2 00:00:10,543 --> 00:00:15,019 which works the same as it does for lists and dictionaries - it clears the set, deleting 3 00:00:15,019 --> 00:00:20,619 all the items in it. I've created a new Python file called remove_items_set, 4 00:00:20,619 --> 00:00:24,660 for these examples. Rather than typing in a load of numbers, we 5 00:00:24,660 --> 00:00:30,113 can pass a range to the set function. I'll type the code to do that, and print the result: 6 00:00:45,973 --> 00:00:50,920 Run the program. We get a set containing all the integers from 7 00:00:50,920 --> 00:00:56,333 zero to 20, inclusive. The values appear in order, but don't be fooled! 8 00:00:56,333 --> 00:01:00,126 Sets have no ordering. Python has to print them out in some order, 9 00:01:00,126 --> 00:01:04,471 and will sometimes display a set of integers in order. But don't think that means the set 10 00:01:04,471 --> 00:01:08,604 is ordered – it isn't. This is an effect of the way sets are implemented, 11 00:01:08,604 --> 00:01:14,523 in CPython. If you're using a different implementation of Python, you may get different output. 12 00:01:14,523 --> 00:01:18,435 I've deliberately used strings in the previous examples, because they make it clear that 13 00:01:18,435 --> 00:01:22,070 the sets are unordered. When you have a set of integers, it can be 14 00:01:22,070 --> 00:01:24,317 less obvious that there's no order to the set. 15 00:01:24,317 --> 00:01:32,499 So let's check out the documentation about sets. I'll go to that page in my browser. 16 00:01:36,281 --> 00:01:42,031 The first sentence makes it clear: "A set object is an unordered collection of distinct hashable objects". 17 00:01:42,031 --> 00:01:46,766 A particular set might print out in order, but it's still an unordered collection. 18 00:01:46,766 --> 00:01:51,168 If you write code that relies on items being in order, then your code will break! 19 00:01:51,168 --> 00:01:56,399 Alright, I wanted to clear up any confusion about that. Let's go back to our code. 20 00:01:56,399 --> 00:02:01,509 Like lists and dictionaries, Python sets have a clear method. It removes all the items in 21 00:02:01,509 --> 00:02:07,249 the set, leaving an empty set. I'll type the code to clear the set, and print the result: 22 00:02:18,785 --> 00:02:20,589 Run the program: 23 00:02:21,821 --> 00:02:26,389 As expected, we get an empty set. Notice that the output isn't an opening and 24 00:02:26,389 --> 00:02:31,860 closing pair of curly braces. That would look like a dictionary, so Python 3.9 prints set 25 00:02:31,860 --> 00:02:36,592 and parentheses. Different versions may display curly braces instead. 26 00:02:36,592 --> 00:02:40,700 There's not much else to say about clear. If you want to clear out all the items in 27 00:02:40,700 --> 00:02:44,610 a list, a dictionary or a set, call the clear method. 28 00:02:44,610 --> 00:02:49,510 I'll comment out those 2 lines, because we need some items in our set. 29 00:02:49,510 --> 00:02:54,419 To delete individual items, you can use the remove or discard methods. 30 00:02:54,419 --> 00:02:59,290 They both do the same thing – they remove an item from the set. The difference is that 31 00:02:59,290 --> 00:03:01,813 remove will raise an exception if the item isn't present. 32 00:03:02,150 --> 00:03:07,557 Some examples will demonstrate the difference. I'll type the code after those lines I commented out: 33 00:03:27,398 --> 00:03:33,121 Ok, run the program, and 10 and 11 are no longer in the set. 34 00:03:33,121 --> 00:03:38,686 Let's see what happens if we discard an item that doesn't exist. I'll add the code to the end of our program: 35 00:03:52,200 --> 00:03:55,190 Run the program: 36 00:03:55,190 --> 00:04:01,211 That works fine. 99 wasn't in the set, so calling discard doesn't change anything. 37 00:04:01,211 --> 00:04:05,582 That's not the case if we use remove. I'll type that code next: 38 00:04:19,507 --> 00:04:25,000 This time, when I run the program, it crashes with an error on line 14. 39 00:04:25,894 --> 00:04:30,720 As the Zen of Python says, "There should be one – and preferably only one – obvious 40 00:04:30,720 --> 00:04:34,710 way to do it". So why do we have both discard and remove? 41 00:04:34,710 --> 00:04:36,590 We've just seen the difference in their behaviour, 42 00:04:36,590 --> 00:04:39,810 and it is an important difference. 43 00:04:41,027 --> 00:04:46,249 There are two different things we might want to do, and having two different methods reflects that: 44 00:04:46,249 --> 00:04:52,636 1: Ensure that, when we're done, something no longer exists. In this case, we don't care 45 00:04:52,636 --> 00:04:55,440 if it was already there or not – we just want it gone. 46 00:04:55,440 --> 00:05:00,448 2: Remove something if it exists, but provide some sort of notification if it doesn't. 47 00:05:00,448 --> 00:05:05,082 These are different use cases, and Python provides an obvious way to do each one; 48 00:05:05,082 --> 00:05:09,460 discard for the first case, and remove for the second. 49 00:05:10,999 --> 00:05:16,570 Airports restrict the items that you can carry in hand luggage. If you're very well organised, you might run 50 00:05:16,570 --> 00:05:21,139 your packing list through a Python program, before packing. You might have the following 51 00:05:21,139 --> 00:05:28,121 set of things that you can't take: Restricted items equals gun, catapult, blade, scissors. 52 00:05:28,121 --> 00:05:31,536 That's not the full list, but those will do for our example. 53 00:05:31,536 --> 00:05:34,037 So your Python program will iterate over that set, 54 00:05:34,037 --> 00:05:38,259 and remove anything it finds from your set of items to pack. 55 00:05:39,783 --> 00:05:44,719 The code might look something like this: For items in restricted items; packing set 56 00:05:44,719 --> 00:05:49,648 dot discard, item. We loop through the items in the set of restricted items, 57 00:05:49,648 --> 00:05:52,950 and discard any restricted items from our packing set. 58 00:05:52,950 --> 00:05:58,030 Using discard there is fine. If you haven't included the item in your packing set, discard 59 00:05:58,030 --> 00:06:04,572 won't raise an exception. The code just moves on to the next item. 60 00:06:04,572 --> 00:06:09,280 Using remove wouldn't be a good idea. When it failed to find "catapult" in the list of 61 00:06:09,280 --> 00:06:12,393 things you intend to pack, the code would crash. 62 00:06:12,393 --> 00:06:15,830 This is an example of when you don't care if something is in the set or not – you 63 00:06:15,830 --> 00:06:20,389 just want to make sure it's not there, after the code has executed. 64 00:06:21,796 --> 00:06:27,564 Lets assume we've created a list of sets. The sets contain the medications that patients are taking, 65 00:06:27,681 --> 00:06:31,706 and our intention is to replace the blood thinning drug warfarin with edoxaban. 66 00:06:31,706 --> 00:06:36,120 Both drugs thin the blood to prevent blood clots, but edoxaban has some advantages over 67 00:06:36,120 --> 00:06:39,050 warfarin. I'm going to assume that we've already identified 68 00:06:39,050 --> 00:06:42,356 those patients who would benefit from changing their medication. 69 00:06:42,356 --> 00:06:47,370 And yes, this is a bit contrived . But it also does represent a real situation that 70 00:06:47,370 --> 00:06:50,720 could arise. So our code has to remove warfarin from the 71 00:06:50,720 --> 00:06:53,988 set of medications that each patient is taking. 72 00:06:55,645 --> 00:07:02,317 The aim here, is to replace one anti-coagulant drug with another. That means we can't just use discard. 73 00:07:02,317 --> 00:07:07,189 When we come to remove warfarin, it's important that the patient was actually taking it. 74 00:07:08,948 --> 00:07:14,639 In this case, it's better that our code crashes. Otherwise, we could end up prescribing the 75 00:07:14,639 --> 00:07:19,450 anti-coagulant drug, edoxaban, to someone who didn't need an anti-coagulant. 76 00:07:19,450 --> 00:07:23,879 remove is a better option here. If we attempt to remove warfarin, and the 77 00:07:23,879 --> 00:07:27,710 patient wasn't taking it, that's a situation we need to be alerted to. 78 00:07:27,710 --> 00:07:32,100 When we use remove, we'll get an error, if warfarin isn't in the set of medications that 79 00:07:32,100 --> 00:07:34,622 the patient was taking. 80 00:07:36,366 --> 00:07:41,899 Now I'm not suggesting that you write code that you know will crash. That's not the point of all this. 81 00:07:41,899 --> 00:07:49,249 The point is that remove will let us know, if we try to delete something that doesn't exist. discard won't. 82 00:07:50,656 --> 00:07:53,009 Later, we'll see how to handle the exception 83 00:07:53,009 --> 00:07:56,869 that remove causes, when you try to remove something that doesn't exist. 84 00:07:56,869 --> 00:08:00,690 At the moment, we're looking at why there are two different methods to delete an item 85 00:08:00,690 --> 00:08:03,409 from a set. So that's the reason why. 86 00:08:03,409 --> 00:08:08,550 In the next two videos, we'll look at the code for both these methods. Examples will 87 00:08:08,550 --> 00:08:12,434 make the difference clear, and give an idea of when you should use each method. 88 00:08:12,434 --> 00:08:15,380 I'll see you in the next video.