Would you like to inspect the original subtitles? These are the user uploaded subtitles that are being translated: 1 00:00:00.05 --> 00:00:03.05 - [Instructor] Once all of the design criteria has been set, 2 00:00:03.05 --> 00:00:06.05 we can start generating options by running the study. 3 00:00:06.05 --> 00:00:09.00 So let's do that in this lesson while learning how 4 00:00:09.00 --> 00:00:12.04 the Generation settings impact the results. 5 00:00:12.04 --> 00:00:14.03 The Generation settings are found down here 6 00:00:14.03 --> 00:00:15.09 at the bottom of the dialog, 7 00:00:15.09 --> 00:00:18.02 within the Generation settings tab. 8 00:00:18.02 --> 00:00:20.06 These settings will change depending on whether 9 00:00:20.06 --> 00:00:24.02 your method is optimize, randomize, cross product, 10 00:00:24.02 --> 00:00:26.00 or like this. 11 00:00:26.00 --> 00:00:29.01 Like this and randomize have the same options, 12 00:00:29.01 --> 00:00:31.06 that is, they have a number of solutions 13 00:00:31.06 --> 00:00:33.04 and they seed input. 14 00:00:33.04 --> 00:00:35.09 Number of solutions lets us specify 15 00:00:35.09 --> 00:00:38.09 how many solutions the study will produce. 16 00:00:38.09 --> 00:00:42.01 So for example, if we're using the like this method, 17 00:00:42.01 --> 00:00:44.08 we have 20 solutions as the default, 18 00:00:44.08 --> 00:00:48.01 which means we would see 20 results. 19 00:00:48.01 --> 00:00:50.00 The same goes for randomize, 20 00:00:50.00 --> 00:00:53.00 which would produce 20 random results 21 00:00:53.00 --> 00:00:55.03 if we were to leave this as 20. 22 00:00:55.03 --> 00:00:58.07 The seed input specifies a starting point for the algorithm 23 00:00:58.07 --> 00:01:00.00 to commence from. 24 00:01:00.00 --> 00:01:02.09 This would relate to a random set of variables 25 00:01:02.09 --> 00:01:05.06 based on the inputs that we've set. 26 00:01:05.06 --> 00:01:08.00 So if we were to leave this at one, 27 00:01:08.00 --> 00:01:10.06 it would produce 20 random solutions. 28 00:01:10.06 --> 00:01:12.03 If we were to run it again, 29 00:01:12.03 --> 00:01:15.02 it would produce another 20 random solutions, 30 00:01:15.02 --> 00:01:17.08 which would most likely be the same as the first 31 00:01:17.08 --> 00:01:21.09 as the seed input or starting point is the same. 32 00:01:21.09 --> 00:01:24.07 If the seed input would change to two however, 33 00:01:24.07 --> 00:01:28.04 on the second run, this would most likely produce 34 00:01:28.04 --> 00:01:32.04 20 different results as the starting point has changed. 35 00:01:32.04 --> 00:01:35.04 Adjusting the seed input can be useful if the initial 36 00:01:35.04 --> 00:01:37.06 20 solutions don't produce the results 37 00:01:37.06 --> 00:01:39.03 that we're after. 38 00:01:39.03 --> 00:01:41.02 And now let's leave it as one. 39 00:01:41.02 --> 00:01:43.04 The cross-product method does not require 40 00:01:43.04 --> 00:01:45.03 any generation settings. 41 00:01:45.03 --> 00:01:48.06 And that is because the number of solutions is determined 42 00:01:48.06 --> 00:01:52.00 by these inputs that we have for the variables. 43 00:01:52.00 --> 00:01:54.08 Because we have five variations for the first 44 00:01:54.08 --> 00:01:57.00 and five variations for the second, 45 00:01:57.00 --> 00:01:59.08 it means we'll get 25 solutions. 46 00:01:59.08 --> 00:02:02.06 The optimize settings are a little bit different 47 00:02:02.06 --> 00:02:07.07 in that we will have a population size and generations. 48 00:02:07.07 --> 00:02:11.02 This is because they utilize genetic algorithms to run. 49 00:02:11.02 --> 00:02:14.05 Population size determines how big the starting 50 00:02:14.05 --> 00:02:16.09 and subsequent population will be, 51 00:02:16.09 --> 00:02:20.05 that is how many possible solutions it will contain. 52 00:02:20.05 --> 00:02:24.03 And generations determines how long the genetic algorithm 53 00:02:24.03 --> 00:02:27.02 will run, that is how many times the population 54 00:02:27.02 --> 00:02:29.07 will evolve in an effort to improve 55 00:02:29.07 --> 00:02:32.04 the fittest design solutions. 56 00:02:32.04 --> 00:02:36.01 For most occasions, the default settings should be fine. 57 00:02:36.01 --> 00:02:38.07 However, let's get ours to run a little bit longer 58 00:02:38.07 --> 00:02:41.07 by changing the Generations to 10. 59 00:02:41.07 --> 00:02:44.09 This means that our population of 12 solutions 60 00:02:44.09 --> 00:02:49.03 will evolve 10 times to try and find the fitter solution. 61 00:02:49.03 --> 00:02:53.01 Keep in mind that it might help to tweak these parameters. 62 00:02:53.01 --> 00:02:56.06 However, sometimes it can have adverse effects. 63 00:02:56.06 --> 00:02:59.02 If the population size is too big 64 00:02:59.02 --> 00:03:01.03 and the generation's too small, 65 00:03:01.03 --> 00:03:04.08 it might not have the time to find the fitter solution. 66 00:03:04.08 --> 00:03:07.00 Or if you have too many generations, 67 00:03:07.00 --> 00:03:09.01 it might run for too long. 68 00:03:09.01 --> 00:03:12.01 Now that we've input those, our study is ready to run. 69 00:03:12.01 --> 00:03:14.06 Before we do that however, let's quickly run over 70 00:03:14.06 --> 00:03:18.01 the last section, which is named issues. 71 00:03:18.01 --> 00:03:21.03 This is very useful to assess before we start our study, 72 00:03:21.03 --> 00:03:24.06 as it displays any errors our inputs might have. 73 00:03:24.06 --> 00:03:28.03 However, if there is an error such as our seed being zero, 74 00:03:28.03 --> 00:03:31.02 for example, it won't let us run the study 75 00:03:31.02 --> 00:03:33.08 and we can see that because the generate button 76 00:03:33.08 --> 00:03:35.04 has been grayed out. 77 00:03:35.04 --> 00:03:38.06 When there is an error, it will show us what that error is 78 00:03:38.06 --> 00:03:40.08 and give us some pretty clear instruction 79 00:03:40.08 --> 00:03:42.08 as to how to fix it. 80 00:03:42.08 --> 00:03:46.07 In our case, let's put our seed input back to one. 81 00:03:46.07 --> 00:03:49.04 This is something to keep in mind when we're running 82 00:03:49.04 --> 00:03:51.03 our studies to make sure that there's nothing 83 00:03:51.03 --> 00:03:52.05 that's going wrong. 84 00:03:52.05 --> 00:03:54.05 As we can say, we have no issues, 85 00:03:54.05 --> 00:03:56.08 so let's go ahead and start generating results 86 00:03:56.08 --> 00:04:00.02 by hitting the generate button. 87 00:04:00.02 --> 00:04:03.07 This brings up the loading screen, which you can see here. 88 00:04:03.07 --> 00:04:05.08 Depending on how complex our study is, 89 00:04:05.08 --> 00:04:08.03 this could take some time. 90 00:04:08.03 --> 00:04:10.07 We'll know when it's finished because the progress counter 91 00:04:10.07 --> 00:04:13.04 on the left will reach its goal 92 00:04:13.04 --> 00:04:15.04 and a green tick will appear. 93 00:04:15.04 --> 00:04:18.02 When it is finished, we get this outcomes page 94 00:04:18.02 --> 00:04:20.00 displaying all of the results. 95 00:04:20.00 --> 00:04:22.07 So next, we'll look at how to decipher the results 96 00:04:22.07 --> 00:04:24.06 and implement them into our project. 7917