WEBVTT 1 00:00:00.620 --> 00:00:02.220 Hi guys and welcome back. 2 00:00:02.220 --> 00:00:05.420 In this video we're going to learn about unpacking arguments 3 00:00:05.420 --> 00:00:07.360 and this is a little bit more complicated 4 00:00:07.360 --> 00:00:10.263 than what we've seen so far, so pleased do bear with me. 5 00:00:11.100 --> 00:00:13.660 I'm going to define a multiply function 6 00:00:13.660 --> 00:00:16.680 that takes in any number of arguments. 7 00:00:16.680 --> 00:00:17.513 Here's how. 8 00:00:19.060 --> 00:00:21.900 We create the multiply function as normal, 9 00:00:21.900 --> 00:00:26.123 and in here, we're going to put, star, args. 10 00:00:27.030 --> 00:00:31.040 Now the star is the important part, the args 11 00:00:31.040 --> 00:00:33.580 is just a variable name that will be used 12 00:00:33.580 --> 00:00:36.290 to collect the arguments. 13 00:00:36.290 --> 00:00:38.573 So let's print the args out. 14 00:00:39.730 --> 00:00:41.101 For now it's not going to multiply anything, 15 00:00:41.101 --> 00:00:42.560 it's just going to print them out, 16 00:00:42.560 --> 00:00:44.910 so what's happening here is we've defined a function 17 00:00:44.910 --> 00:00:47.340 and we've said, that the function has 18 00:00:48.780 --> 00:00:53.550 a set of arguments that will be collected into 19 00:00:53.550 --> 00:00:57.720 a tuple of arguments when the function gets called. 20 00:00:57.720 --> 00:01:02.010 So if we do multiply and then we do one, three and five, 21 00:01:02.010 --> 00:01:06.185 notice that here I'm passing different arguments. 22 00:01:06.185 --> 00:01:09.303 But this only has seemingly a single parameter. 23 00:01:09.303 --> 00:01:10.890 So I'm passing three different arguments 24 00:01:10.890 --> 00:01:14.870 and what's gonna happen is they are going to be collected 25 00:01:14.870 --> 00:01:19.870 into this args variable, so I'm gonna save this and run it. 26 00:01:19.950 --> 00:01:22.010 And you can see that what we get printed out 27 00:01:22.010 --> 00:01:25.410 when we run the function is a tuple of my arguments, 28 00:01:25.410 --> 00:01:28.423 so we've got one, three and five printed out as a tuple. 29 00:01:29.370 --> 00:01:32.470 If we wanted to multiply all these numbers together, 30 00:01:32.470 --> 00:01:36.990 we can just iterate over the arguments so for arg in args 31 00:01:36.990 --> 00:01:40.750 we can say total, equal total times arg, 32 00:01:40.750 --> 00:01:45.710 and then return total, then we can print that out. 33 00:01:45.710 --> 00:01:47.330 Now what we're doing here is we're starting off 34 00:01:47.330 --> 00:01:49.710 with a variable which is one and then we're multiplying 35 00:01:49.710 --> 00:01:52.380 that variable each of the arguments, 36 00:01:52.380 --> 00:01:54.330 and then we're returning the final value. 37 00:01:54.330 --> 00:01:59.180 If we run that we get an error because we left the S there 38 00:01:59.180 --> 00:02:01.330 so let's run that again. 39 00:02:01.330 --> 00:02:04.180 There we go we get 15 printed out. 40 00:02:04.180 --> 00:02:08.030 So this is a way collecting multiple arguments 41 00:02:08.030 --> 00:02:11.590 into a single variable when you are calling a function. 42 00:02:11.590 --> 00:02:14.730 And this has a number of uses that you will definitely 43 00:02:14.730 --> 00:02:18.210 be seeing quite a lot as you programme in Python. 44 00:02:18.210 --> 00:02:21.560 Notice that you can call this function with one argument 45 00:02:21.560 --> 00:02:23.980 if you want, and that's totally fine as well, 46 00:02:23.980 --> 00:02:27.170 it just creates a tuple of a single element. 47 00:02:27.170 --> 00:02:29.710 Let's go into another example let's say we have 48 00:02:29.710 --> 00:02:33.190 an add function with two parameters x and y, 49 00:02:33.190 --> 00:02:36.660 and then you have a list of numbers, three and five. 50 00:02:36.660 --> 00:02:40.560 So just as we can use the asterisk to collect arguments 51 00:02:40.560 --> 00:02:44.170 into one parameter, we can go the other way round 52 00:02:44.170 --> 00:02:47.750 and we can use the asterisk to de-structure arguments 53 00:02:47.750 --> 00:02:49.340 into multiple parameters. 54 00:02:49.340 --> 00:02:53.480 So you can do add, star nums, and what this'll do 55 00:02:53.480 --> 00:02:58.240 is it will use three for x and five for y 56 00:02:58.240 --> 00:03:01.550 so it is de-structuring the nums variable 57 00:03:01.550 --> 00:03:04.060 and splitting it out into multiple values 58 00:03:04.060 --> 00:03:07.367 such that one value can go to one parameter. 59 00:03:07.367 --> 00:03:10.170 Notice that if you do just nums here, 60 00:03:10.170 --> 00:03:13.820 you're going to pass three and five, the list itself 61 00:03:13.820 --> 00:03:17.310 as the value for x and y is not going to have a value 62 00:03:17.310 --> 00:03:18.950 so you're going to get an error. 63 00:03:18.950 --> 00:03:23.950 So by doing this it passes one value for each parameter. 64 00:03:23.980 --> 00:03:26.570 Of course you do need to have the same number of values 65 00:03:26.570 --> 00:03:29.210 in this list, as parameters in this function, 66 00:03:29.210 --> 00:03:30.610 otherwise you're going to have an error 67 00:03:30.610 --> 00:03:32.030 where you're gonna have too many parameters 68 00:03:32.030 --> 00:03:33.654 or too many arguments. 69 00:03:33.654 --> 00:03:37.130 Let's print this out so you can check that it works. 70 00:03:37.130 --> 00:03:39.270 There you have it, eight comes out. 71 00:03:39.270 --> 00:03:43.160 Now this can also work with named arguments 72 00:03:43.160 --> 00:03:45.690 so remember you can call the add function 73 00:03:45.690 --> 00:03:49.530 by doing x equals three and y equal five. 74 00:03:49.530 --> 00:03:53.270 So you can do something like that using these asterisks. 75 00:03:53.270 --> 00:03:55.840 Let's say that you've got your nums and these are now 76 00:03:55.840 --> 00:04:00.840 a dictionary with a value for x, and a value for y. 77 00:04:00.940 --> 00:04:04.740 If you wanted to use this dictionary as the argument 78 00:04:04.740 --> 00:04:09.420 to these functions, you would have to do nums x 79 00:04:09.420 --> 00:04:13.700 and nums y, like that, which is kind of ugly 80 00:04:13.700 --> 00:04:15.770 and also difficult to read, you've started to get a lot 81 00:04:15.770 --> 00:04:17.470 of brackets and square brackets and so on. 82 00:04:17.470 --> 00:04:19.940 If you wanted to be even more specific you can do 83 00:04:19.940 --> 00:04:24.580 x equals nums x and y equal nums y. 84 00:04:24.580 --> 00:04:27.740 And here is where the magic happens. 85 00:04:27.740 --> 00:04:29.630 You can see that the parameter name 86 00:04:30.660 --> 00:04:35.170 is the same as the key in the dictionary that we are using 87 00:04:35.170 --> 00:04:40.170 as a value for that parameter, so x is the key x 88 00:04:40.470 --> 00:04:43.220 and y is the key y. 89 00:04:43.220 --> 00:04:47.200 Instead of doing this entire thing, we can tell Python 90 00:04:47.200 --> 00:04:52.200 to pass in this dictionary as named arguments, 91 00:04:52.630 --> 00:04:54.970 where the argument will be equal to the value 92 00:04:54.970 --> 00:04:58.750 associated with that same name for a key. 93 00:04:58.750 --> 00:05:02.240 So essentially doing exactly this, but just do 94 00:05:02.240 --> 00:05:05.940 start star, nums, and that will do the same thing. 95 00:05:05.940 --> 00:05:08.240 So because we're using start star, it will say 96 00:05:08.240 --> 00:05:11.020 okay you've got a dictionary with two keys, 97 00:05:11.020 --> 00:05:13.160 what I'm gonna do is I'm gonna pass in each key 98 00:05:13.160 --> 00:05:16.300 as a named argument, and the value is going to be 99 00:05:16.300 --> 00:05:18.590 that associated with the key, so it'll say 100 00:05:18.590 --> 00:05:21.760 x equals nums x and y equals nums y. 101 00:05:21.760 --> 00:05:24.220 I know that's quite complicated thing, 102 00:05:24.220 --> 00:05:26.760 but it simplifies your code so much 103 00:05:26.760 --> 00:05:29.370 when you can use dictionaries like that. 104 00:05:29.370 --> 00:05:31.960 Let's go back to our multiply function 105 00:05:31.960 --> 00:05:35.750 that we had earlier on with this collected set of arguments. 106 00:05:35.750 --> 00:05:38.310 And I'm going to create anther function called apply 107 00:05:38.310 --> 00:05:42.463 and now we're going to say start args, comma operator. 108 00:05:43.340 --> 00:05:46.030 So this is a special bit of syntax in Python 109 00:05:46.030 --> 00:05:49.930 and what it means is collect all the positional arguments 110 00:05:49.930 --> 00:05:54.510 into this tuple args, and also have 111 00:05:54.510 --> 00:05:57.780 a named argument at the end. 112 00:05:57.780 --> 00:06:02.710 This creates a compulsory or required named argument. 113 00:06:02.710 --> 00:06:04.640 So when you're calling this function you're gonna pass in 114 00:06:04.640 --> 00:06:07.610 as many arguments as you want and they will all be collected 115 00:06:07.610 --> 00:06:11.120 into args and at the end you must pass in 116 00:06:11.120 --> 00:06:13.350 a named argument for this one. 117 00:06:13.350 --> 00:06:16.160 So let me just give it a body there, and we're gonna say 118 00:06:16.160 --> 00:06:18.620 if the operator is equal to the asterisk, 119 00:06:18.620 --> 00:06:21.130 we're going to return multiply of args, 120 00:06:21.130 --> 00:06:23.940 otherwise if the operator is equal to plus, 121 00:06:23.940 --> 00:06:25.620 we're going to return the sum of args 122 00:06:25.620 --> 00:06:27.660 and finally we're going to return no valid operator 123 00:06:27.660 --> 00:06:28.810 if it's neither of them. 124 00:06:28.810 --> 00:06:30.790 Now I have a bug in this function 125 00:06:30.790 --> 00:06:33.300 and you may notice what it is based on 126 00:06:33.300 --> 00:06:35.260 what we've seen already, but I know that you don't have 127 00:06:35.260 --> 00:06:38.430 a lot of practise with this yet so it is possible 128 00:06:38.430 --> 00:06:40.650 to not notice it, that's fine. 129 00:06:40.650 --> 00:06:44.890 We're going to print apply of one three six and seven, 130 00:06:44.890 --> 00:06:47.660 and the operator is going to be plus. 131 00:06:47.660 --> 00:06:49.780 So this is how this function must be called. 132 00:06:49.780 --> 00:06:52.290 You can pass in as many arguments as you want. 133 00:06:52.290 --> 00:06:54.043 They will be collected by args. 134 00:06:54.043 --> 00:06:57.710 Finally you must pass in a named argument 135 00:06:57.710 --> 00:06:59.180 for this one there. 136 00:06:59.180 --> 00:07:02.030 If you just pass in plus, Python's gonna try to 137 00:07:02.030 --> 00:07:05.438 collect it with args, so you're gonna miss this parameter. 138 00:07:05.438 --> 00:07:09.560 Indeed if we run this, you'll see that you get an error. 139 00:07:09.560 --> 00:07:12.590 Missing one required keyword only argument. 140 00:07:12.590 --> 00:07:15.310 So you do need the operator in there. 141 00:07:15.310 --> 00:07:18.700 So if we run this, it'll all work, you get 17 back 142 00:07:18.700 --> 00:07:21.230 because that's the addition of all of these numbers. 143 00:07:21.230 --> 00:07:24.833 However if you change this to a asterisk, 144 00:07:25.880 --> 00:07:30.570 then you get back a tuple, so why this tuple? 145 00:07:30.570 --> 00:07:33.850 Why not the multiplication of all the numbers? 146 00:07:33.850 --> 00:07:37.030 Is there a bug in our multiply function? 147 00:07:37.030 --> 00:07:39.520 Well you might be tempted to go through 148 00:07:39.520 --> 00:07:41.860 and see what the multiply function is doing. 149 00:07:41.860 --> 00:07:45.040 Multiply function starts with a variable called total 150 00:07:45.040 --> 00:07:47.220 that has the value one, and then it goes through 151 00:07:47.220 --> 00:07:50.560 the arguments and makes total equal to 152 00:07:50.560 --> 00:07:53.010 total multiplied by the arg. 153 00:07:53.010 --> 00:07:56.500 So how on earth is it managing to end up 154 00:07:56.500 --> 00:07:58.580 with a tuple being returned? 155 00:07:58.580 --> 00:08:02.260 Well if you multiply by a tuple, 156 00:08:02.260 --> 00:08:04.650 you actually end up with that tuple. 157 00:08:04.650 --> 00:08:09.650 So that means that args must be a tuple of tuples 158 00:08:10.260 --> 00:08:15.260 such that the first argument, the first arg in the tuple 159 00:08:15.272 --> 00:08:18.680 is itself a tuple, such that you're multiplying 160 00:08:18.680 --> 00:08:23.130 the tuple by one, and ending up with the tuple itself 161 00:08:23.130 --> 00:08:26.600 and then returning that, so the best way to do this 162 00:08:26.600 --> 00:08:28.410 is to use a debugger, but we haven't learned 163 00:08:28.410 --> 00:08:29.790 how to use a debugger yet. 164 00:08:29.790 --> 00:08:31.860 So we're gonna print the argumetns and you'll see that 165 00:08:31.860 --> 00:08:33.830 you're gonna get something a little bit weird. 166 00:08:33.830 --> 00:08:37.650 When you print the args, you get a tuple of tuples. 167 00:08:37.650 --> 00:08:39.870 And then when you iterate over the args, 168 00:08:39.870 --> 00:08:42.520 arg is the inner tuple. 169 00:08:42.520 --> 00:08:45.951 You're multiplying that by one, making it equal to the total 170 00:08:45.951 --> 00:08:49.570 such that total is the tuple, and then because that's 171 00:08:49.570 --> 00:08:53.150 the only value in args, that is the end of the for loop 172 00:08:53.150 --> 00:08:54.910 and you return the tuple. 173 00:08:54.910 --> 00:08:59.000 The reason for that is because we did not pass in 174 00:08:59.000 --> 00:09:01.580 individual arguments to the multiply function, 175 00:09:01.580 --> 00:09:02.963 we passed in a tuple. 176 00:09:04.010 --> 00:09:06.540 We want to put a star here. 177 00:09:06.540 --> 00:09:08.310 The reason for that is because when we call 178 00:09:08.310 --> 00:09:11.540 the apply function we are passing in four different 179 00:09:11.540 --> 00:09:13.430 arguments that are being collected 180 00:09:13.430 --> 00:09:16.850 into one tuple, by this star. 181 00:09:16.850 --> 00:09:19.330 But when we call the multiply function, 182 00:09:19.330 --> 00:09:22.360 we wanna pass in four different arguments 183 00:09:22.360 --> 00:09:25.630 so that they can be collected by args, 184 00:09:25.630 --> 00:09:29.410 inside the multiplier function and then used in here. 185 00:09:29.410 --> 00:09:32.460 But before, we were passing the args tuple itself, 186 00:09:32.460 --> 00:09:34.590 the collected tuple in this function. 187 00:09:34.590 --> 00:09:38.350 What we wanna do is we wanna unpack it into four values 188 00:09:38.350 --> 00:09:41.590 and pass them as individual values to this function 189 00:09:41.590 --> 00:09:44.920 so that they can be re-collected by args here. 190 00:09:44.920 --> 00:09:47.530 The other option is to remove the star 191 00:09:47.530 --> 00:09:49.970 so that you are passing a tuple directly 192 00:09:49.970 --> 00:09:52.940 to the multiplier function, and then moving this star 193 00:09:52.940 --> 00:09:56.150 so that the multiply function won't try to collect 194 00:09:56.150 --> 00:09:59.520 all its positional arguments into another tuple. 195 00:09:59.520 --> 00:10:02.740 However I do like passing multiple different arguments 196 00:10:02.740 --> 00:10:04.500 so I'm going to keep this stars there, 197 00:10:04.500 --> 00:10:05.820 but that's a choice for you to make, 198 00:10:05.820 --> 00:10:09.160 you can either pass arguments, or you can pass a tuple. 199 00:10:09.160 --> 00:10:11.830 That is a very common problem and quite a difficult 200 00:10:11.830 --> 00:10:13.890 to diagnose problem as well. 201 00:10:13.890 --> 00:10:15.690 So I just wanted to go through it so that 202 00:10:15.690 --> 00:10:17.400 if you encounter that in the future, 203 00:10:17.400 --> 00:10:18.510 you'll know what's going on. 204 00:10:18.510 --> 00:10:19.880 Thanks for joining me in this video. 205 00:10:19.880 --> 00:10:22.070 Hope it's been useful and not too confusing. 206 00:10:22.070 --> 00:10:23.720 And I'll see you in the next one.