1 00:00:04,740 --> 00:00:06,839 In this video, we're going to have a look 2 00:00:06,839 --> 00:00:09,539 at how the any and all functions can be 3 00:00:09,539 --> 00:00:12,209 useful, when used with a comprehension 4 00:00:12,209 --> 00:00:15,299 or generator expression. So let's start by 5 00:00:15,299 --> 00:00:20,730 creating a new python file, and we'll 6 00:00:20,730 --> 00:00:22,320 go with calling this one 7 00:00:22,320 --> 00:00:30,600 anycomprehension.py. So for the first 8 00:00:30,600 --> 00:00:32,549 example, let's say we're writing an email 9 00:00:32,549 --> 00:00:34,949 client in Python. I'm going to start on 10 00:00:34,949 --> 00:00:37,460 the first line - I'm typing from data 11 00:00:37,460 --> 00:00:43,320 import people basic plants underscore list, or 12 00:00:43,320 --> 00:00:45,210 underscore plants underscore list, and as 13 00:00:45,210 --> 00:00:48,870 well as plants underscore list. So as I 14 00:00:48,870 --> 00:00:50,790 mentioned, we're going to assume we're 15 00:00:50,790 --> 00:00:52,800 writing an email client in Python. We've 16 00:00:52,800 --> 00:00:54,420 allowed the user to choose the people 17 00:00:54,420 --> 00:00:56,730 who receive the email from a contacts 18 00:00:56,730 --> 00:00:58,980 database, maybe. The user wants to send 19 00:00:58,980 --> 00:01:01,020 the email and the first thing we might 20 00:01:01,020 --> 00:01:03,239 want to do is check that all the people 21 00:01:03,239 --> 00:01:05,280 have an email address. If they do we can 22 00:01:05,280 --> 00:01:07,320 attempt to send the email. If they don't 23 00:01:07,320 --> 00:01:09,450 we let the user edit the list of 24 00:01:09,450 --> 00:01:11,640 recipients. To do that we need to check 25 00:01:11,640 --> 00:01:13,860 the second field of each of the people. 26 00:01:13,860 --> 00:01:15,810 We can create a list containing that 27 00:01:15,810 --> 00:01:18,659 field by using a list comprehension. So 28 00:01:18,659 --> 00:01:19,830 let's go ahead and do that. So on line 29 00:01:19,830 --> 00:01:22,939 3 I'm going to type if all 30 00:01:22,939 --> 00:01:26,580 parentheses square bracket person 31 00:01:26,580 --> 00:01:28,710 left square bracket 1 right 32 00:01:28,710 --> 00:01:30,750 square bracket space and its for 33 00:01:30,750 --> 00:01:34,830 person in people, then we're closing 34 00:01:34,830 --> 00:01:38,520 right square bracket there, right 35 00:01:38,520 --> 00:01:42,439 parenthesis and then a colon. Then print 36 00:01:42,439 --> 00:01:46,820 parentheses double quote Sending email, 37 00:01:46,820 --> 00:01:49,049 closing double quote closing parenthesis. 38 00:01:49,049 --> 00:01:50,970 Then on the next line, we'll go back into 39 00:01:50,970 --> 00:01:54,450 an else colon, and print on the line 40 00:01:54,450 --> 00:01:57,240 after that, double quotes in parentheses 41 00:01:57,240 --> 00:02:03,320 User must edit the list of recipients, 42 00:02:03,320 --> 00:02:06,689 double quote right curly brace, or right 43 00:02:06,689 --> 00:02:08,579 parenthesis, I should say. So the 44 00:02:08,579 --> 00:02:11,790 all function will return true if all of the 45 00:02:11,790 --> 00:02:14,129 email addresses aren't empty, false if one 46 00:02:14,129 --> 00:02:16,290 or more of them are empty. So if we run 47 00:02:16,290 --> 00:02:18,340 the program now, 48 00:02:18,340 --> 00:02:20,810 you can see the user must edit the list 49 00:02:20,810 --> 00:02:23,060 of recipients, and that's exactly what we 50 00:02:23,060 --> 00:02:26,120 want, or returns false because a 51 00:02:26,120 --> 00:02:27,560 couple of the people are missing an email 52 00:02:27,560 --> 00:02:30,650 address. If we go back to our data dot 53 00:02:30,650 --> 00:02:37,790 py, and you can see that Eric Idle on 54 00:02:37,790 --> 00:02:39,920 line 6 and Michael Palin on line 9 55 00:02:39,920 --> 00:02:41,689 haven't got email addresses, which 56 00:02:41,689 --> 00:02:44,239 is consistent with our code failing, 57 00:02:44,239 --> 00:02:46,280 or returning false I should say, for that 58 00:02:46,280 --> 00:02:49,219 reason. So as you can see, using all is 59 00:02:49,219 --> 00:02:51,319 much simpler than iterating over the 60 00:02:51,319 --> 00:02:53,930 list, checking each email address. Another 61 00:02:53,930 --> 00:02:56,090 advantage of all, and any too for that 62 00:02:56,090 --> 00:02:58,189 matter, is that they short-circuit the 63 00:02:58,189 --> 00:03:01,099 evaluation of the iterable. As soon as 64 00:03:01,099 --> 00:03:02,090 the result is known, 65 00:03:02,090 --> 00:03:04,189 they stop checking any more values. In 66 00:03:04,189 --> 00:03:06,949 the case of all, it can return false as 67 00:03:06,949 --> 00:03:09,260 soon as the false value is found. In the 68 00:03:09,260 --> 00:03:11,510 case of any, it can return true as soon 69 00:03:11,510 --> 00:03:13,090 as a true value is found. 70 00:03:13,090 --> 00:03:15,590 Remember that gotcha from a few videos 71 00:03:15,590 --> 00:03:18,560 ago, though? The code will fail if the 72 00:03:18,560 --> 00:03:20,750 list is empty, and we can see that by 73 00:03:20,750 --> 00:03:22,669 setting it to empty before using it. 74 00:03:22,669 --> 00:03:24,229 So we come back up to here, and on line 75 00:03:24,229 --> 00:03:28,939 3, I type people equals and left 76 00:03:28,939 --> 00:03:30,199 and right square brackets to create an empty 77 00:03:30,199 --> 00:03:34,629 list, and if we run this code again now, 78 00:03:34,629 --> 00:03:41,180 Sending email, as you can see there. 79 00:03:41,180 --> 00:03:42,769 And by the way, just ignore this warning over 80 00:03:42,769 --> 00:03:45,439 here: Redeclared 'people' defined above 81 00:03:45,439 --> 00:03:48,109 without usage. We're only 82 00:03:48,109 --> 00:03:50,209 changing people here to save modifying 83 00:03:50,209 --> 00:03:53,060 data.py, but again, you can see 84 00:03:53,060 --> 00:03:54,859 that wasn't what we wanted when we had an 85 00:03:54,859 --> 00:03:57,829 empty list here, defined. Our email 86 00:03:57,829 --> 00:03:59,509 program would attempt to send the email 87 00:03:59,509 --> 00:04:01,370 to an empty list of recipients and clearly, 88 00:04:01,370 --> 00:04:03,409 that's not what we'd want to do. So we 89 00:04:03,409 --> 00:04:05,329 either need a separate test or we can 90 00:04:05,329 --> 00:04:08,299 combine them, using the bool function. So 91 00:04:08,299 --> 00:04:10,400 if I modify line 5, and in fact, change 92 00:04:10,400 --> 00:04:14,060 that to if - and we'll have a test at the 93 00:04:14,060 --> 00:04:20,000 start first - if bool person and, and then 94 00:04:20,000 --> 00:04:21,769 all and as we had it before, 95 00:04:21,769 --> 00:04:24,080 so basically, adding that second test. 96 00:04:24,080 --> 00:04:28,020 That should've been people not person, 97 00:04:28,020 --> 00:04:32,550 and if we run this again now, "User must 98 00:04:32,550 --> 00:04:35,759 edit the list of recipients. So that works 99 00:04:35,759 --> 00:04:37,889 by adding a separate test, and just to 100 00:04:37,889 --> 00:04:39,750 confirm if that works now, we can 101 00:04:39,750 --> 00:04:42,509 actually comment out this line 3, when we 102 00:04:42,509 --> 00:04:43,919 created an empty list and run it again, 103 00:04:43,919 --> 00:04:47,220 and that still works with the actual 104 00:04:47,220 --> 00:04:49,680 valid data. So always test what are 105 00:04:49,680 --> 00:04:52,259 called edge cases. Things like values 106 00:04:52,259 --> 00:04:54,539 being 0 or negative, lists and 107 00:04:54,539 --> 00:04:56,550 dictionaries being empty, values that are 108 00:04:56,550 --> 00:04:58,289 far greater than you'd expect, things like 109 00:04:58,289 --> 00:05:00,900 that. Making sure that your code handles 110 00:05:00,900 --> 00:05:02,849 invalid data is the difference 111 00:05:02,849 --> 00:05:06,270 between an OK program and a great one. Alright, 112 00:05:06,270 --> 00:05:08,970 so for an example of using any, let's assume 113 00:05:08,970 --> 00:05:11,190 we're working on a shopping site that 114 00:05:11,190 --> 00:05:14,370 sells selections of plant seeds. The 115 00:05:14,370 --> 00:05:16,470 site's created some variety packs to 116 00:05:16,470 --> 00:05:19,590 make it easier for customers to choose. So 117 00:05:19,590 --> 00:05:20,789 instead of clicking on loads of 118 00:05:20,789 --> 00:05:22,860 individual seed packets, the site makes 119 00:05:22,860 --> 00:05:24,599 it easy for them to just specify some 120 00:05:24,599 --> 00:05:26,819 likes and dislikes, and offer some packs 121 00:05:26,819 --> 00:05:28,620 that fit what they want. So one of the 122 00:05:28,620 --> 00:05:30,960 options is grass. If you're a gardener, 123 00:05:30,960 --> 00:05:32,550 you'll know that there's some lovely 124 00:05:32,550 --> 00:05:34,830 grasses that can enhance a garden, but 125 00:05:34,830 --> 00:05:36,330 some people don't want to plant grasses. 126 00:05:36,330 --> 00:05:38,610 So what we can do here is, we can use any 127 00:05:38,610 --> 00:05:40,919 to check if the plants in a variety 128 00:05:40,919 --> 00:05:43,229 pack contain at least one plant that's a 129 00:05:43,229 --> 00:05:45,330 grass. So I'm going to add this code 130 00:05:45,330 --> 00:05:47,870 after the last example. 131 00:05:47,870 --> 00:05:52,610 So on line 10, I can type if any 132 00:05:52,610 --> 00:05:57,150 parentheses square bracket plant dot 133 00:05:57,150 --> 00:06:02,599 plant underscore type is equal to, two equals 134 00:06:02,599 --> 00:06:06,180 double quotes grass and then a space 135 00:06:06,180 --> 00:06:10,919 there, then for plant in plants 136 00:06:10,919 --> 00:06:13,229 underscore list. Then we're going to have right 137 00:06:13,229 --> 00:06:15,389 square bracket right parenthesis and a 138 00:06:15,389 --> 00:06:18,479 colon. Then we can type print parentheses 139 00:06:18,479 --> 00:06:22,319 double quotes, this pack, This pack 140 00:06:22,319 --> 00:06:25,620 contains grass, closing double quote and 141 00:06:25,620 --> 00:06:28,500 parenthesis. Then introduce an else 142 00:06:28,500 --> 00:06:32,069 and in the else case we'll actually print no 143 00:06:32,069 --> 00:06:37,380 grasses in this pack and close it off. So 144 00:06:37,380 --> 00:06:41,600 when we run that now, 145 00:06:41,600 --> 00:06:44,580 you can see we've got this message down the 146 00:06:44,580 --> 00:06:45,600 bottom left hand corner, "This pack 147 00:06:45,600 --> 00:06:47,970 contains grass". So we can see that this 148 00:06:47,970 --> 00:06:50,250 pack does in fact contain at least 149 00:06:50,250 --> 00:06:52,979 one grass type. The comprehension in this 150 00:06:52,979 --> 00:06:54,870 example, creates a list containing true 151 00:06:54,870 --> 00:06:56,759 or false, depending on whether the 152 00:06:56,759 --> 00:06:59,130 condition's true for each plant in the 153 00:06:59,130 --> 00:07:02,160 list. Using a name tuple, I think and I'm 154 00:07:02,160 --> 00:07:03,449 sure you'll agree, makes the code 155 00:07:03,449 --> 00:07:05,789 more readable and saves us having to check 156 00:07:05,789 --> 00:07:08,130 the index position of plant underscore 157 00:07:08,130 --> 00:07:10,800 type in the tuple. Alright, so you can 158 00:07:10,800 --> 00:07:11,970 do the same thing with a 159 00:07:11,970 --> 00:07:14,039 dictionary instead and that's going to make a good 160 00:07:14,039 --> 00:07:15,690 challenge. Now just before I actually 161 00:07:15,690 --> 00:07:17,550 start the challenge though, the challenge 162 00:07:17,550 --> 00:07:18,990 is, it's going to involve a dictionary, 163 00:07:18,990 --> 00:07:21,150 as I mentioned, so I need 164 00:07:21,150 --> 00:07:22,650 to change the definitions up here. So I'm 165 00:07:22,650 --> 00:07:23,610 going to do that, I'm going to delete 166 00:07:23,610 --> 00:07:25,470 that basic underscore plants underscore 167 00:07:25,470 --> 00:07:27,330 list and instead what we're going to 168 00:07:27,330 --> 00:07:30,479 do is import as the third import, our 169 00:07:30,479 --> 00:07:31,979 plants underscore dict because we'll 170 00:07:31,979 --> 00:07:35,070 need that for the challenge. Alright so 171 00:07:35,070 --> 00:07:37,020 let's look at the challenge. The challenge is 172 00:07:37,020 --> 00:07:38,849 to use any and a comprehension, or 173 00:07:38,849 --> 00:07:41,250 generator expression if you prefer, to 174 00:07:41,250 --> 00:07:43,169 check the plants in plants underscore 175 00:07:43,169 --> 00:07:45,270 dict to see if there are any grasses in 176 00:07:45,270 --> 00:07:46,919 there. So you want to do that first and 177 00:07:46,919 --> 00:07:48,389 then run your code again after that, 178 00:07:48,389 --> 00:07:50,760 searching for cactus to test that it 179 00:07:50,760 --> 00:07:53,550 still works when there aren't any. 180 00:07:53,550 --> 00:07:54,930 Alright so that's the challenge. Pause the 181 00:07:54,930 --> 00:07:56,400 video. When you're ready to come back 182 00:07:56,400 --> 00:07:58,289 through my solution, come back and we'll 183 00:07:58,289 --> 00:08:01,510 go through that, so pause the video now. 184 00:08:01,510 --> 00:08:03,250 Alright so welcome back. 185 00:08:03,250 --> 00:08:05,230 Hopefully, you managed to get that solved. Let's go 186 00:08:05,230 --> 00:08:06,700 through and I'm going to show you the 187 00:08:06,700 --> 00:08:08,830 solution, and I'll add that just after 188 00:08:08,830 --> 00:08:10,360 the examples we've been working 189 00:08:10,360 --> 00:08:12,880 through in this video. I'm going to start on line 190 00:08:12,880 --> 00:08:17,880 15 by typing if any parentheses plant 191 00:08:17,880 --> 00:08:21,160 dot plant underscore type, that's equals 192 00:08:21,160 --> 00:08:25,120 to double quotes grass, capital G, for 193 00:08:25,120 --> 00:08:31,210 plant in plants underscore dict dot 194 00:08:31,210 --> 00:08:34,510 values then parentheses and a right 195 00:08:34,510 --> 00:08:36,539 parenthesis to close it off and a colon. 196 00:08:36,539 --> 00:08:39,690 So that's the case - we're going to print 197 00:08:39,690 --> 00:08:42,690 parentheses double quotes This dict 198 00:08:42,690 --> 00:08:47,740 contains grasses, else colon on the next 199 00:08:47,740 --> 00:08:49,410 line, print on the line after that, 200 00:08:49,410 --> 00:08:52,990 parentheses double quotes No grasses 201 00:08:52,990 --> 00:08:56,890 in the dict. So what I've done there, I've used 202 00:08:56,890 --> 00:08:58,770 a generator expression to save memory. 203 00:08:58,770 --> 00:09:00,640 Because we're dealing with a dictionary, 204 00:09:00,640 --> 00:09:02,980 the generator expression iterates over 205 00:09:02,980 --> 00:09:05,740 the dictionaries values view, as you can 206 00:09:05,740 --> 00:09:06,430 see over here 207 00:09:06,430 --> 00:09:08,950 on line 15, and checks the plant 208 00:09:08,950 --> 00:09:11,740 underscore type field of each value. So 209 00:09:11,740 --> 00:09:13,560 if we run that to make sure it works, 210 00:09:13,560 --> 00:09:15,670 and we can see that the dict contains 211 00:09:15,670 --> 00:09:18,070 grasses. Now you may have iterated over 212 00:09:18,070 --> 00:09:20,560 the keys instead, and that would also work 213 00:09:20,560 --> 00:09:22,090 fine - that's another way of doing it. 214 00:09:22,090 --> 00:09:23,770 So let's have a look at how to do 215 00:09:23,770 --> 00:09:25,990 that, and there we'll start, we'll 216 00:09:25,990 --> 00:09:28,930 start on line 20; if any plants 217 00:09:28,930 --> 00:09:32,650 underscore dict left square bracket key 218 00:09:32,650 --> 00:09:35,350 right square bracket dot plant 219 00:09:35,350 --> 00:09:39,210 underscore type, that's equal to Grass, 220 00:09:39,210 --> 00:09:44,500 and then for key in plants underscore 221 00:09:44,500 --> 00:09:46,060 dict. So that's the alternative way 222 00:09:46,060 --> 00:09:47,980 of doing it. Again iterating over the 223 00:09:47,980 --> 00:09:49,600 keys, and everything else will, 224 00:09:49,600 --> 00:09:51,130 effectively, be the same after that. We 225 00:09:51,130 --> 00:09:52,780 can just copy the same three lines there, 226 00:09:52,780 --> 00:10:00,700 and if we run that, we get the same 227 00:10:00,700 --> 00:10:04,450 output, as you can see there. So the code's 228 00:10:04,450 --> 00:10:06,430 a little bit less readable, but may 229 00:10:06,430 --> 00:10:08,500 perform better than iterating over the 230 00:10:08,500 --> 00:10:11,320 values view. If you, if you think that's 231 00:10:11,320 --> 00:10:12,670 likely to be an issue in your particular 232 00:10:12,670 --> 00:10:15,100 application, you can test the performance 233 00:10:15,100 --> 00:10:17,470 of both methods. We saw how to do that, of 234 00:10:17,470 --> 00:10:18,700 course, in an earlier video in this 235 00:10:18,700 --> 00:10:20,770 section of the course. Alright, so lastly, 236 00:10:20,770 --> 00:10:22,450 let's change the last generator 237 00:10:22,450 --> 00:10:24,730 expression to check for cactus instead, 238 00:10:24,730 --> 00:10:26,580 just to make sure that it's working 239 00:10:26,580 --> 00:10:29,710 and there is what we're looking for, and also 240 00:10:29,710 --> 00:10:31,480 when there isn't. So we can go ahead and 241 00:10:31,480 --> 00:10:40,210 just change Grass there to Cactus. 242 00:10:40,210 --> 00:10:49,870 Change the output, run the program, and you 243 00:10:49,870 --> 00:10:52,600 can see there's now cactii in the dict. 244 00:10:52,600 --> 00:10:54,310 Remember to test if there's an empty 245 00:10:54,310 --> 00:10:56,500 dictionary as well. It does work - 246 00:10:56,500 --> 00:10:58,420 any behaves as expected with an 247 00:10:58,420 --> 00:11:00,550 empty interable, but really you should 248 00:11:00,550 --> 00:11:02,950 still test it anyway. Alright, so that's 249 00:11:02,950 --> 00:11:05,620 the any and all functions. They're very 250 00:11:05,620 --> 00:11:07,350 easy to use and can be extremely 251 00:11:07,350 --> 00:11:09,340 useful. So I'm going to finish this section 252 00:11:09,340 --> 00:11:11,770 now with a quick look at lambdas. Now 253 00:11:11,770 --> 00:11:13,330 that's something else that Guido doesn't 254 00:11:13,330 --> 00:11:15,760 like but they can have their uses. So 255 00:11:15,760 --> 00:11:19,110 I'll see you in the next video.