1 00:00:00,210 --> 00:00:07,560 Hello and welcome to this lecture, where we are going to solve the problem related to the transport 2 00:00:07,560 --> 00:00:12,000 of products using M. L Rose Library. 3 00:00:12,300 --> 00:00:20,100 This is another interesting library where you can find implementations of genetic algorithms. 4 00:00:20,400 --> 00:00:21,820 Follow this link! 5 00:00:21,960 --> 00:00:31,440 If you want to know more detail, you can see some informations about optimization problems and differently 6 00:00:31,440 --> 00:00:36,540 from deep, which is focused on evolutionary algorithms. 7 00:00:36,930 --> 00:00:43,920 These order library is focused on optimization algorithms. 8 00:00:44,370 --> 00:00:53,420 For example, we can click algorithms here and then our implementations of the hillclimb algorithm. 9 00:00:53,880 --> 00:01:01,230 We can see here the simulated annealing and also genetic algorithms. 10 00:01:01,740 --> 00:01:09,420 They are different kinds of algorithms to minimize or maximize of value. 11 00:01:09,690 --> 00:01:19,470 So instead of using genetic algorithms to solve this problem, you can also use simulated annealing 12 00:01:19,470 --> 00:01:22,110 and also the hillclimb. 13 00:01:22,590 --> 00:01:27,570 The good point is that you have easy access to the source codes. 14 00:01:27,900 --> 00:01:37,650 For example, you can click Sirs here, and you can take a look at the genetic algorithm implementation. 15 00:01:37,940 --> 00:01:48,120 Maybe if you was to make some changes to our algorithm implemented from scratch, you can see more information 16 00:01:48,120 --> 00:01:48,540 here. 17 00:01:49,200 --> 00:01:51,840 Now, let's install the library. 18 00:01:52,380 --> 00:01:54,420 Let's type PIP. 19 00:01:54,870 --> 00:01:58,000 Install Email Rose. 20 00:01:58,650 --> 00:02:06,840 Run this sale of codes we are using version one point three point zero. 21 00:02:07,110 --> 00:02:12,450 If you have some problem in music, you would think the codes we are going to implement. 22 00:02:12,720 --> 00:02:21,020 Maybe it is because of different versions so you can type B install. 23 00:02:21,610 --> 00:02:33,570 Email rose equal one point three point zero, so the codes will be compatible with the one we are going 24 00:02:33,570 --> 00:02:35,190 to implement now. 25 00:02:35,490 --> 00:02:42,140 The next step we need to impart to the library in parts and URLs. 26 00:02:42,480 --> 00:02:43,980 Run this cell. 27 00:02:44,250 --> 00:02:51,330 No error was returns, so it means the library was successfully installed. 28 00:02:51,750 --> 00:02:55,530 Now we need to define the list of products. 29 00:02:55,890 --> 00:03:00,720 Any stands of defining use in the four months of a class. 30 00:03:01,020 --> 00:03:07,920 We can also paste this code, which is a simple list in Python. 31 00:03:08,280 --> 00:03:13,980 In each one of the positions, we can see the features about the products. 32 00:03:14,310 --> 00:03:19,380 The name this base and also the price you can use. 33 00:03:19,380 --> 00:03:26,250 The class we implement as before, and also you can define in the form of a list. 34 00:03:26,640 --> 00:03:28,740 We can create this variable. 35 00:03:29,280 --> 00:03:31,770 Lets type products. 36 00:03:33,090 --> 00:03:37,290 You can see all the information about our data. 37 00:03:37,470 --> 00:03:46,530 Let's also create another variable limit equals three, which means the available space in the truck. 38 00:03:46,860 --> 00:03:54,360 Now in these should define the in function again and choose not to be so repetitive. 39 00:03:54,720 --> 00:04:00,060 I will just based this code to here fitness function. 40 00:04:00,420 --> 00:04:06,270 We sans a solution which is composed of zeros and ones. 41 00:04:06,570 --> 00:04:14,760 Just a reminder, zero means that we are not going to load the products on the truck, and one means 42 00:04:14,760 --> 00:04:18,510 that we are going to load the product on the truck. 43 00:04:18,840 --> 00:04:29,460 At the end, we return the cost and we don't need to type a call here because it is not necessary. 44 00:04:29,490 --> 00:04:36,990 Using this library, let's create the function and now we can create a new variable. 45 00:04:37,650 --> 00:04:48,630 Fitness equals email rose custom fitness because we are using our own fitness function. 46 00:04:49,230 --> 00:04:51,170 Let's just dive here. 47 00:04:51,540 --> 00:04:53,080 Fit in this function. 48 00:04:53,220 --> 00:04:55,500 This same name of the function? 49 00:04:56,130 --> 00:04:57,630 Run this code. 50 00:04:58,140 --> 00:04:59,770 Now we needs to refer. 51 00:04:59,890 --> 00:05:01,540 Zandt's, the problem? 52 00:05:01,930 --> 00:05:06,700 Email rules got this crap. 53 00:05:07,000 --> 00:05:14,710 Obesity this crats means that we are going to work only with integer numbers. 54 00:05:15,130 --> 00:05:19,210 We need to specify the length of this solution. 55 00:05:19,610 --> 00:05:24,400 Land products we can type this code to here. 56 00:05:25,030 --> 00:05:30,070 Land products just correct this typo. 57 00:05:30,070 --> 00:05:39,880 Here we can see that the results equals 14, which means the amount of products in our data. 58 00:05:40,330 --> 00:05:49,360 Then, when these two is specified, this fit in this function, which will be equal to fitness. 59 00:05:49,930 --> 00:06:00,580 This variable we define the higher than the X parameter maximize equals true because we want higher 60 00:06:00,580 --> 00:06:01,450 values. 61 00:06:01,990 --> 00:06:06,340 And finally, Max Vo equals chu. 62 00:06:06,610 --> 00:06:13,420 It means we was to generate only zeros and ones. 63 00:06:13,690 --> 00:06:25,000 So when we define number two, the library will generate only zeros and once we run this sale of codes. 64 00:06:26,080 --> 00:06:30,310 And finally, we can create two new variables. 65 00:06:30,610 --> 00:06:43,030 The first one, the best solution and the best fit in this are the best value equals email rules we 66 00:06:43,030 --> 00:06:46,300 can call the genetic algorithm. 67 00:06:47,080 --> 00:06:49,540 We will sans the problem. 68 00:06:50,110 --> 00:06:54,700 Then the experimenter pops size, population size. 69 00:06:54,940 --> 00:07:04,840 For example, we can define only 20 individuals, and there is another parameter mutational prob the 70 00:07:04,840 --> 00:07:09,220 probability of mutation zero point zero one. 71 00:07:09,520 --> 00:07:18,580 This same parameters according to our last implementation, and we don't need to specify the number 72 00:07:18,580 --> 00:07:28,960 of generations because the algorithm itself will try to find the optimal number of generations after 73 00:07:28,960 --> 00:07:30,160 ruining this code. 74 00:07:30,370 --> 00:07:39,460 We can take a look at the best solution and the best fit in this let's run this code. 75 00:07:39,880 --> 00:07:43,290 We can see the chromosome, and here it is. 76 00:07:43,290 --> 00:07:48,880 Solution twenty four thousand three hundred and eighty five. 77 00:07:49,300 --> 00:07:58,270 Every time we run this code, we will get a different value because the initialization of the chromosome 78 00:07:58,480 --> 00:08:01,630 in the beginning of the algorithm is random. 79 00:08:02,080 --> 00:08:11,440 Let's run these codes a couple of times, and we can see results quite similar to the one we got before. 80 00:08:11,890 --> 00:08:21,970 To show the products that we are going challenge, we can implement the far loop for I in range length 81 00:08:22,450 --> 00:08:24,370 based solution. 82 00:08:24,820 --> 00:08:30,010 We are going through these lists of zeros and ones. 83 00:08:30,700 --> 00:08:37,450 If the best solution position, I equals one. 84 00:08:38,200 --> 00:08:42,820 Let's just sprints the name of the product. 85 00:08:43,240 --> 00:08:48,070 We can access our list products position. 86 00:08:48,070 --> 00:09:02,320 I am position zero and then the price of the product lets access again our list products position. 87 00:09:02,320 --> 00:09:05,590 I am position to just two. 88 00:09:05,590 --> 00:09:16,670 Badru understands we can type where products we can see all information if we was to get the cell phone. 89 00:09:17,410 --> 00:09:21,610 It is in position zero and position one. 90 00:09:22,000 --> 00:09:27,520 We can type a products position when we get only the cell phone. 91 00:09:27,940 --> 00:09:32,860 If we want to get the name, it is in position zero. 92 00:09:33,490 --> 00:09:37,570 For this reason, we have these implementation here. 93 00:09:37,930 --> 00:09:46,690 We can also access the price products product one and position two. 94 00:09:48,410 --> 00:09:53,060 Ran this code, and we have the price of the product. 95 00:09:53,390 --> 00:09:58,520 Now we can run this code, and here is our results. 96 00:09:58,970 --> 00:10:03,830 This is the list of products that we are going to load on the drug. 97 00:10:04,190 --> 00:10:13,610 As you could see, it's much easier to work with Melrose Library in the next lecture and we will present 98 00:10:13,610 --> 00:10:13,910 you. 99 00:10:13,940 --> 00:10:21,530 This second case is the the the second problem that we are going to solve using both libraries. 100 00:10:21,860 --> 00:10:22,550 See you there!