1 00:00:04,730 --> 00:00:07,910 When dealing with data in tuples, you can 2 00:00:07,910 --> 00:00:09,410 make the code more readable by 3 00:00:09,410 --> 00:00:12,469 using named tuples. In this video, we're going 4 00:00:12,469 --> 00:00:13,490 to have a quick look at these named 5 00:00:13,490 --> 00:00:15,590 tuples, what they are and also how 6 00:00:15,590 --> 00:00:17,779 to create them. Now there's some good 7 00:00:17,779 --> 00:00:19,340 documentation, if we swing over to the 8 00:00:19,340 --> 00:00:22,730 Python website. We've got 9 00:00:22,730 --> 00:00:24,440 some information here relating to named 10 00:00:24,440 --> 00:00:28,130 tuples, and this will be a good thing 11 00:00:28,130 --> 00:00:29,689 for you to read because it does explain 12 00:00:29,689 --> 00:00:31,609 them well, so I suggest you check that out 13 00:00:31,609 --> 00:00:33,769 after you've been through this video. 14 00:00:33,769 --> 00:00:34,940 One thing I would like to point out, 15 00:00:34,940 --> 00:00:37,550 though, is that just before the change history 16 00:00:37,550 --> 00:00:39,620 in example down here, this comment here, 17 00:00:39,620 --> 00:00:41,210 and it talks about the fact that named 18 00:00:41,210 --> 00:00:43,699 tuples are lightweight and require no 19 00:00:43,699 --> 00:00:45,769 more memory than regular tuples. And that 20 00:00:45,769 --> 00:00:47,330 is certainly true. They do need a bit 21 00:00:47,330 --> 00:00:50,479 more work to create, but not a lot. 22 00:00:50,479 --> 00:00:52,489 So let's see how to use them in a 23 00:00:52,489 --> 00:00:54,620 Python program. So you should have downloaded 24 00:00:54,620 --> 00:00:56,659 data.py from the resources 25 00:00:56,659 --> 00:00:58,790 section of the last video. So go ahead and 26 00:00:58,790 --> 00:01:00,309 grab that if you haven't got that already, 27 00:01:00,309 --> 00:01:03,920 and we'll go back to finder, in my case 28 00:01:03,920 --> 00:01:06,560 on a Mac. We'll go to Windows Explorer 29 00:01:06,560 --> 00:01:07,880 if you're on a Windows machine. Take a copy 30 00:01:07,880 --> 00:01:10,759 of that downloaded file, swing back 31 00:01:10,759 --> 00:01:12,740 to IntelliJ and we're going to right 32 00:01:12,740 --> 00:01:16,579 click and paste to bring that file into 33 00:01:16,579 --> 00:01:19,670 Python. And in this video, we're going 34 00:01:19,670 --> 00:01:23,539 to be using plant based tuples, named tuples. 35 00:01:23,539 --> 00:01:25,340 You can see examples of those on 36 00:01:25,340 --> 00:01:27,799 lines 15 and 16, but there is also 37 00:01:27,799 --> 00:01:29,960 a people list at the start there, on line 3. 38 00:01:29,960 --> 00:01:31,719 We'll be using that in the next video. 39 00:01:31,719 --> 00:01:34,820 So how do we actually use named 40 00:01:34,820 --> 00:01:36,950 tuples? Well firstly, as you can see, 41 00:01:36,950 --> 00:01:39,590 on line 1 here in our data.py, we're 42 00:01:39,590 --> 00:01:41,149 importing named tuple from the 43 00:01:41,149 --> 00:01:43,310 collections module. So that lets us 44 00:01:43,310 --> 00:01:44,960 create named tuples, as you've 45 00:01:44,960 --> 00:01:45,590 probably guessed. 46 00:01:45,590 --> 00:01:48,920 Now before creating a named tuple, we have 47 00:01:48,920 --> 00:01:51,020 to define what the names are, and again 48 00:01:51,020 --> 00:01:53,390 you can see in lines 15 and 16 49 00:01:53,390 --> 00:01:56,359 we've defined two named tuples; plant and plant 50 00:01:56,359 --> 00:01:59,420 details. The first argument to named tuple 51 00:01:59,420 --> 00:02:02,539 is the type name as a string. I can't see 52 00:02:02,539 --> 00:02:03,920 any reason to make that different to the 53 00:02:03,920 --> 00:02:06,350 name that we assign the result to, but 54 00:02:06,350 --> 00:02:07,159 you can if you want, 55 00:02:07,159 --> 00:02:09,800 if causing confusion is your thing. 56 00:02:09,800 --> 00:02:12,500 So I'll change the first plant on line 15 57 00:02:12,500 --> 00:02:17,030 to P, and if we scroll down and have a 58 00:02:17,030 --> 00:02:17,920 look, 59 00:02:17,920 --> 00:02:20,170 notice on line 42 down here, we've now 60 00:02:20,170 --> 00:02:22,180 got errors where I've used plant, and 61 00:02:22,180 --> 00:02:24,700 that confirms it's the returned object 62 00:02:24,700 --> 00:02:27,160 that we need to use when creating our 63 00:02:27,160 --> 00:02:29,530 named tuples. Because if you go back now 64 00:02:29,530 --> 00:02:31,540 to line 15 again, if we undo that change, 65 00:02:31,540 --> 00:02:36,130 and instead change the second plant to P, 66 00:02:36,130 --> 00:02:40,680 this one here, let's scroll down and have a look, 67 00:02:40,680 --> 00:02:43,690 we're not getting any errors. So you can 68 00:02:43,690 --> 00:02:45,070 do that, you can make them different if 69 00:02:45,070 --> 00:02:45,520 you want, 70 00:02:45,520 --> 00:02:47,650 again if causing confusion is your thing, 71 00:02:47,650 --> 00:02:49,090 I suggest you go ahead and do that, but 72 00:02:49,090 --> 00:02:52,030 seriously, I can't see any reason to make 73 00:02:52,030 --> 00:02:53,380 those different. So generally, they'll be 74 00:02:53,380 --> 00:02:55,090 the same. So what we've done here then, 75 00:02:55,090 --> 00:02:57,550 on line 15, we've created the new name 76 00:02:57,550 --> 00:03:01,480 tuple using plant but their type is P. 77 00:03:01,480 --> 00:03:03,519 Now to see why that's confusing, 78 00:03:03,519 --> 00:03:05,890 I won't undo that change just yet. 79 00:03:05,890 --> 00:03:07,900 What we'll do is create a new Python file 80 00:03:07,900 --> 00:03:11,709 in the project, and we call this one named 81 00:03:11,709 --> 00:03:21,100 tupletest, and we'll go ahead and type 82 00:03:21,100 --> 00:03:23,940 some code in here; so line one, from data 83 00:03:23,940 --> 00:03:27,760 import basic_plants underscore 84 00:03:27,760 --> 00:03:31,600 list, comma then plants_list. 85 00:03:31,600 --> 00:03:32,799 On line three, we're gonna type 86 00:03:32,799 --> 00:03:36,370 print parentheses plants underscore list left 87 00:03:36,370 --> 00:03:37,780 square bracket zero right square 88 00:03:37,780 --> 00:03:40,180 bracket, then closing parentheses. So I put the 89 00:03:40,180 --> 00:03:42,160 data in data.py to save having 90 00:03:42,160 --> 00:03:44,440 to type it in again, and also so that we 91 00:03:44,440 --> 00:03:46,299 could reuse it in different example 92 00:03:46,299 --> 00:03:48,730 programs. So the code, as you can see, we 93 00:03:48,730 --> 00:03:50,130 start by importing the two tuples, 94 00:03:50,130 --> 00:03:52,120 basic_plants underscore 95 00:03:52,120 --> 00:03:53,920 list is a list of ordinary tuples, 96 00:03:53,920 --> 00:03:55,959 and plant_list is a list of 97 00:03:55,959 --> 00:04:02,790 named tuples. So if we run the program, 98 00:04:02,790 --> 00:04:05,560 you can see down the bottom, it shows the 99 00:04:05,560 --> 00:04:08,440 first item as being of type P. 100 00:04:08,440 --> 00:04:10,660 So everything works. It's just confusing to 101 00:04:10,660 --> 00:04:13,269 create something of type P when we use 102 00:04:13,269 --> 00:04:16,029 plant to create it. So don't do it 103 00:04:16,029 --> 00:04:17,620 unless you've got a really good reason to do so. 104 00:04:17,620 --> 00:04:20,440 I'm gonna switch back to data.py 105 00:04:20,440 --> 00:04:22,919 and I'm just going to undo that change now. OK. 106 00:04:22,919 --> 00:04:25,960 Alright, so that was the first 107 00:04:25,960 --> 00:04:28,210 argument. Now the second argument is 108 00:04:28,210 --> 00:04:30,640 just a list of the names we want to use for 109 00:04:30,640 --> 00:04:31,720 each of the fields 110 00:04:31,720 --> 00:04:34,570 in the named tuple. Now these names must 111 00:04:34,570 --> 00:04:37,000 conform to the rules for Python variable 112 00:04:37,000 --> 00:04:39,670 names, with the added restriction that 113 00:04:39,670 --> 00:04:41,320 they can't start with an underscore, 114 00:04:41,320 --> 00:04:43,180 and that's mentioned in the documentation we 115 00:04:43,180 --> 00:04:45,130 looked at earlier. So in this case, on 116 00:04:45,130 --> 00:04:47,350 line 15, our plant named tuple has 117 00:04:47,350 --> 00:04:50,760 four fields: name, scientific_name, 118 00:04:50,760 --> 00:04:53,860 lifecycle and plant_type. 119 00:04:53,860 --> 00:04:56,560 Now the plant details tuple, on line 16, is 120 00:04:56,560 --> 00:04:58,930 the same but it doesn't have a name 121 00:04:58,930 --> 00:05:00,970 field. That's because it's used in a 122 00:05:00,970 --> 00:05:02,620 dictionary and the name is the 123 00:05:02,620 --> 00:05:04,870 dictionary key, and we can see that if we 124 00:05:04,870 --> 00:05:06,610 go down and have a look at that on line 125 00:05:06,610 --> 00:05:10,480 64 onwards. You can see there 126 00:05:10,480 --> 00:05:12,250 the dictionary set up and the name of being 127 00:05:12,250 --> 00:05:15,250 the key. Now getting back to our 128 00:05:15,250 --> 00:05:17,890 definitions on lines 15 and 16, on line 129 00:05:17,890 --> 00:05:20,140 15 specifically, you can see here that 130 00:05:20,140 --> 00:05:21,940 I used a list to specify the field names. 131 00:05:21,940 --> 00:05:25,240 However, Python also allows a space or 132 00:05:25,240 --> 00:05:27,760 comma separated string, and that can be 133 00:05:27,760 --> 00:05:29,380 useful if you're reading the data from a 134 00:05:29,380 --> 00:05:31,780 file, for example. So to see what I mean, I 135 00:05:31,780 --> 00:05:33,730 can change that line now, by removing the 136 00:05:33,730 --> 00:05:40,419 left and right square brackets, as well 137 00:05:40,419 --> 00:05:43,480 as the internal apostrophes. So obviously, 138 00:05:43,480 --> 00:05:45,130 leave the one on the left of name. I'll 139 00:05:45,130 --> 00:05:47,260 leave that as it is there, but delete the 140 00:05:47,260 --> 00:05:54,190 rest of them, and you can see they've 141 00:05:54,190 --> 00:05:56,140 removed all the apostrophes 142 00:05:56,140 --> 00:05:57,760 now, and we're just left with an 143 00:05:57,760 --> 00:05:59,200 apostrophe at the start to the left of 144 00:05:59,200 --> 00:06:02,140 name, and to the right of plant_type. 145 00:06:02,140 --> 00:06:04,600 So if we go back, 146 00:06:04,600 --> 00:06:06,610 now, to namedtupletest.py that we 147 00:06:06,610 --> 00:06:09,820 created, we can print the names from our 148 00:06:09,820 --> 00:06:11,260 basic_plants_list 149 00:06:11,260 --> 00:06:13,870 with a for loop. So let's 150 00:06:13,870 --> 00:06:14,890 go ahead and do that. We'll add that code 151 00:06:14,890 --> 00:06:17,560 from line 5 onwards. So I'm going to type for plant in basic_plant_list. 152 00:06:22,210 --> 00:06:25,210 On the next line, after the colon, print plant zero, 153 00:06:32,229 --> 00:06:36,580 and if you run that, no surprises - we get the names 154 00:06:36,580 --> 00:06:39,700 of the plants printed out, and we can 155 00:06:39,700 --> 00:06:41,050 maybe put this over here on the right 156 00:06:41,050 --> 00:06:43,490 hand side, 157 00:06:43,490 --> 00:06:45,080 just so we can see the names a bit 158 00:06:45,080 --> 00:06:47,539 easier. We could do the same with a named 159 00:06:47,539 --> 00:06:49,970 tuple but we can also use the field 160 00:06:49,970 --> 00:06:52,819 name instead. So to do that we just use 161 00:06:52,819 --> 00:06:55,039 dot notation, just as if it was an 162 00:06:55,039 --> 00:06:58,580 attribute of a class instance. So I can 163 00:06:58,580 --> 00:07:01,220 just comment that code out now, on lines 164 00:07:01,220 --> 00:07:04,729 five and six, and on line eight, we'll 165 00:07:04,729 --> 00:07:08,690 start by typing for plant in plants_list 166 00:07:08,690 --> 00:07:13,610 colon, print, plant and 167 00:07:13,610 --> 00:07:20,000 using dot notation, dot and name. 168 00:07:20,000 --> 00:07:22,009 This makes the code more readable, and we 169 00:07:22,009 --> 00:07:23,090 don't have to remember that the name 170 00:07:23,090 --> 00:07:24,919 appears in position zero of the tuple. 171 00:07:24,919 --> 00:07:28,069 So if you run that, we get the same 172 00:07:28,069 --> 00:07:30,710 output. Now we can still use an index if 173 00:07:30,710 --> 00:07:32,599 you want to, but it really doesn't make a 174 00:07:32,599 --> 00:07:35,659 lot of sense. Plant.scientific_name 175 00:07:35,659 --> 00:07:37,789 will give us the second field but 176 00:07:37,789 --> 00:07:42,830 we could also do plant left square 177 00:07:42,830 --> 00:07:46,909 bracket one right square bracket, run, and 178 00:07:46,909 --> 00:07:48,169 see both fields like that. 179 00:07:48,169 --> 00:07:50,270 You can see we're getting the first two 180 00:07:50,270 --> 00:07:52,370 fields printed there. You wouldn't write 181 00:07:52,370 --> 00:07:54,469 code like this, mixing field names 182 00:07:54,469 --> 00:07:56,870 and indexes, but it can be useful to use the 183 00:07:56,870 --> 00:07:58,759 index sometimes, if it's in a loop 184 00:07:58,759 --> 00:08:01,490 control variable, for example. The point 185 00:08:01,490 --> 00:08:02,960 here though is, we don't lose anything 186 00:08:02,960 --> 00:08:05,599 that can be done with ordinary tuples - 187 00:08:05,599 --> 00:08:07,810 we get extra features with named tuples. 188 00:08:07,810 --> 00:08:10,639 named tuples support all the operations 189 00:08:10,639 --> 00:08:13,130 that you can perform on ordinary tuples. 190 00:08:13,130 --> 00:08:14,780 We can unpack them and iterate over them, 191 00:08:14,780 --> 00:08:17,210 for example, and there's three additional 192 00:08:17,210 --> 00:08:18,800 functions that ordinary tuples don't 193 00:08:18,800 --> 00:08:21,139 have. The documentation covers these well, 194 00:08:21,139 --> 00:08:23,210 but I'll give you an example of the 195 00:08:23,210 --> 00:08:25,490 underscore replace function. Now 196 00:08:25,490 --> 00:08:27,020 the additional functions and two data 197 00:08:27,020 --> 00:08:28,909 attributes have names prefixed 198 00:08:28,909 --> 00:08:31,639 with underscore. Remember, I mentioned earlier 199 00:08:31,639 --> 00:08:33,020 in this video, that you can't use an 200 00:08:33,020 --> 00:08:35,149 underscore before the field names, so 201 00:08:35,149 --> 00:08:36,740 this ensures that there's no conflict 202 00:08:36,740 --> 00:08:39,559 between plant.replace, referring to a 203 00:08:39,559 --> 00:08:41,479 field called replace, and 204 00:08:41,479 --> 00:08:44,420 plant._replace(), referring 205 00:08:44,420 --> 00:08:47,390 to the function. So underscore replace is 206 00:08:47,390 --> 00:08:49,970 a useful way to change the value of one 207 00:08:49,970 --> 00:08:52,339 or more fields in the tuple. Named tuples 208 00:08:52,339 --> 00:08:54,440 are immutable, but you can use the 209 00:08:54,440 --> 00:08:56,360 underscore replace function as a short 210 00:08:56,360 --> 00:08:57,860 and convenient way to 211 00:08:57,860 --> 00:09:00,140 create a new named tuple with some values 212 00:09:00,140 --> 00:09:02,600 changed. So let's have a look how 213 00:09:02,600 --> 00:09:04,670 to go about that. So 214 00:09:04,670 --> 00:09:06,670 on line eleven, I'm just gonna type print 215 00:09:06,670 --> 00:09:10,730 parentheses, space out the output. Then 216 00:09:10,730 --> 00:09:12,790 on line thirteen we'll start with example equals plants_list, then left square bracket, 217 00:09:43,490 --> 00:09:49,130 0, right square bracket, print example. Then on line 15, example equals example._replace, 218 00:09:49,130 --> 00:09:50,930 lifecycle equals Annual, single quote right square bracket. 219 00:09:50,930 --> 00:09:53,920 Then we'll print example. So the Andromeda plant is an evergreen. 220 00:09:53,920 --> 00:09:56,420 That's the value of its lifecycle field, so let's just run that. 221 00:09:56,420 --> 00:09:58,430 Alright, so the Andromeda plant is an evergreen. 222 00:09:58,430 --> 00:10:00,860 That's the value of its lifecycle field. This code prints 223 00:10:00,860 --> 00:10:02,810 that out so that we can check, and then use the 224 00:10:02,810 --> 00:10:05,029 underscore replace function to create a 225 00:10:05,029 --> 00:10:07,459 new named tuple with the 226 00:10:07,459 --> 00:10:10,160 life cycle set to annual. If we go scroll 227 00:10:10,160 --> 00:10:12,860 over, we can see that. There's evergreen 228 00:10:12,860 --> 00:10:14,510 on the first line of output, up here, 229 00:10:14,510 --> 00:10:17,269 and it's changed then on the second line 230 00:10:17,269 --> 00:10:18,980 of output to annual, and all the other 231 00:10:18,980 --> 00:10:20,300 fields, note, are keeping their original 232 00:10:20,300 --> 00:10:23,480 values. So we saw how to do something 233 00:10:23,480 --> 00:10:25,550 similar with ordinary tuples, back in 234 00:10:25,550 --> 00:10:27,980 section seven. This is a useful shortcut 235 00:10:27,980 --> 00:10:29,779 and makes it more obvious what the code 236 00:10:29,779 --> 00:10:32,209 is doing. You may want to use this 237 00:10:32,209 --> 00:10:34,040 method if you've read records from a 238 00:10:34,040 --> 00:10:36,829 database, for example. If the data changes, you can 239 00:10:36,829 --> 00:10:39,260 update the tuple and write the changes 240 00:10:39,260 --> 00:10:41,660 back to the database. Well, that's about 241 00:10:41,660 --> 00:10:43,339 all I've got to say about named tuples. 242 00:10:43,339 --> 00:10:45,980 We'll return to any and all in the next 243 00:10:45,980 --> 00:10:48,620 video, and then use these named tuples in 244 00:10:48,620 --> 00:10:51,079 some of the examples. See you in the next 245 00:10:51,079 --> 00:10:53,290 video.