Would you like to inspect the original subtitles? These are the user uploaded subtitles that are being translated: 1 00:00:00,420 --> 00:00:06,230 Hello everyone and welcome to the lecture on factor in categorical matrices this lecture we're going 2 00:00:06,230 --> 00:00:09,850 to look at how to create factor in categorical matrices in our. 3 00:00:10,050 --> 00:00:13,140 Let's go ahead and jump to our studio to get started. 4 00:00:13,140 --> 00:00:15,300 All right so here here our studio. 5 00:00:15,480 --> 00:00:20,520 And in this lecture we're mainly going to be discussing the factor function and it's used for creating 6 00:00:20,520 --> 00:00:22,210 categorical matrices. 7 00:00:22,470 --> 00:00:27,510 This specific function will become really useful when we begin to apply data analysis and machine learning 8 00:00:27,510 --> 00:00:29,210 techniques to our data. 9 00:00:29,220 --> 00:00:32,160 This is also sometimes and then is creating dummy variables. 10 00:00:32,160 --> 00:00:35,630 So you may be familiar with it in that context. 11 00:00:35,670 --> 00:00:40,230 So let's go to start by show an example of why and how we can build a sort of matrix. 12 00:00:40,390 --> 00:00:50,580 I want to go ahead and create a vector called Animal Nalut just go ahead and pretend that where have 13 00:00:50,580 --> 00:00:52,120 some sort of animal sanctuary. 14 00:00:52,130 --> 00:00:54,040 And there's dogs and cats in it. 15 00:00:54,090 --> 00:00:59,090 So we'll go ahead and just label this the C for cat D for dog. 16 00:00:59,340 --> 00:01:07,180 Let's go ahead and put another dog in there and then we'll start with two more cats. 17 00:01:07,180 --> 00:01:07,870 All right. 18 00:01:08,160 --> 00:01:11,690 So we have our animal vector where C is for cat D is for dog. 19 00:01:11,700 --> 00:01:16,890 And so we're trying to represent some sort of animal sanctuary and let's say each of these specific 20 00:01:16,890 --> 00:01:20,320 animals also has a corresponding ID number. 21 00:01:20,430 --> 00:01:23,440 So we'll go ahead and make an ID number vector. 22 00:01:23,970 --> 00:01:29,130 Know let's just go ahead and say One two three four five. 23 00:01:29,130 --> 00:01:34,440 All right so we have our animals in Adie's and we want to convert the animal vector into information 24 00:01:34,440 --> 00:01:40,260 that an algorithm or equation can understand more easily meaning we want to begin to check how many 25 00:01:40,260 --> 00:01:45,030 categories or factor levels are actually in our character vector. 26 00:01:45,330 --> 00:01:51,600 So we can do is we can pass the vector through the factor function like this. 27 00:01:51,600 --> 00:01:54,230 So we're going to go in and say factor. 28 00:01:55,080 --> 00:02:00,470 And so then notice I'm going to go ahead and pass an animal. 29 00:02:00,810 --> 00:02:07,980 And so what you get back is this vector that also notice that it has these levels attached to it C and 30 00:02:07,980 --> 00:02:08,790 D. 31 00:02:08,850 --> 00:02:16,470 So I'm going to go ahead and assign this factor of animal and I'll assign it to let's just go ahead 32 00:02:16,470 --> 00:02:22,180 and say fact Annie as some variable name for this. 33 00:02:22,320 --> 00:02:24,580 So we're going to save that factor. 34 00:02:24,690 --> 00:02:25,250 All right. 35 00:02:25,410 --> 00:02:29,340 So we can see that we have two levels D-N.C. for dog and cat. 36 00:02:29,610 --> 00:02:34,310 So in our There's going to be two distinct types of categorical variables. 37 00:02:34,380 --> 00:02:38,670 There's going to be ordinal categorical variables in nominal categorical variables. 38 00:02:38,790 --> 00:02:42,780 So the nominal categorical variables let me go in and type this out. 39 00:02:42,780 --> 00:02:44,540 You can also reference the notes for this. 40 00:02:44,770 --> 00:02:50,310 So right now we're talking about nominal must go ahead and make that a little brighter by not putting 41 00:02:50,310 --> 00:02:51,660 it as a comment. 42 00:02:51,660 --> 00:02:55,230 So the nominal categorical variables don't have any order. 43 00:02:55,350 --> 00:02:57,180 So you can imagine that dogs and cats. 44 00:02:57,180 --> 00:02:59,480 There's no order to them. 45 00:02:59,520 --> 00:03:06,570 Well some people may prefer dogs for cats but in reality there's no order to them versus ordinal categorical 46 00:03:06,570 --> 00:03:08,370 variables as the name implies. 47 00:03:08,400 --> 00:03:18,420 Do have some sort of order so we can think of nominal there's no order versus let's go ahead and say 48 00:03:20,160 --> 00:03:24,660 ordinal there is an order. 49 00:03:24,750 --> 00:03:27,690 So nominal an example that is what we have here. 50 00:03:27,840 --> 00:03:34,770 This animal so there's no order versus dogs and cats but something that could be ordinal of order could 51 00:03:34,770 --> 00:03:44,970 be something like temperatures so we could have something like or Dohm cat with let's say cold 52 00:03:48,300 --> 00:03:51,890 medium and hot. 53 00:03:51,960 --> 00:03:59,700 So this is a ordinal categorical variable because even though we're dealing with strings here and different 54 00:03:59,700 --> 00:04:01,080 categories. 55 00:04:01,080 --> 00:04:05,910 So cold is a category medium is a category of temperature hot as category temperature. 56 00:04:05,910 --> 00:04:07,620 You can see that there's an order to them. 57 00:04:07,650 --> 00:04:11,280 It should be cold medium hot versus dogs and cats. 58 00:04:11,310 --> 00:04:14,690 So Dog is a category C is a category for cat. 59 00:04:14,850 --> 00:04:19,470 There's no order of dogs before cats are cast before dogs. 60 00:04:19,470 --> 00:04:20,150 OK. 61 00:04:20,640 --> 00:04:27,120 So if you actually want to assign an order to these variables you would be able to do that while using 62 00:04:27,120 --> 00:04:28,250 the factor function. 63 00:04:28,290 --> 00:04:34,680 And you can actually pass in an argument with ordered equals true and pass in the levels argument as 64 00:04:34,680 --> 00:04:34,990 well. 65 00:04:34,990 --> 00:04:37,510 So we're going to show you how we can do that. 66 00:04:37,560 --> 00:04:47,610 So let's say we have a vector temp's and it's going to be called Medium a high bigger than just copy 67 00:04:47,610 --> 00:04:49,170 and paste this. 68 00:04:49,290 --> 00:04:53,780 And we're also going to put in a couple of more instances so I'll say there's another instance of hot 69 00:04:53,830 --> 00:04:54,320 . 70 00:04:54,690 --> 00:04:58,320 Another instance of Ha we'll say there's another axis of cold. 71 00:04:58,590 --> 00:05:03,440 And finally one more instance of median. 72 00:05:03,820 --> 00:05:04,050 Right. 73 00:05:04,050 --> 00:05:06,820 So we have our attempts vector here. 74 00:05:06,840 --> 00:05:10,520 So it's caught cold medium hot ha hot cold medium. 75 00:05:10,540 --> 00:05:10,740 Right. 76 00:05:10,740 --> 00:05:12,670 So we have these three categories. 77 00:05:12,690 --> 00:05:23,340 Let's go ahead and pass this into a variable using factor and let me just bring this back up by clearing 78 00:05:23,340 --> 00:05:29,110 that I'm going to go ahead and say temp's is that vector of cold medium hot et cetera. 79 00:05:29,130 --> 00:05:39,780 Various instances and I'll say the factored version of this called The Factor function pass in my temps 80 00:05:39,860 --> 00:05:40,850 vector. 81 00:05:41,280 --> 00:05:45,750 I'm going to say ordered is equal to true. 82 00:05:45,780 --> 00:05:50,310 Remember I didn't do this for the dogs and cats vector because there was no order that made sense for 83 00:05:50,310 --> 00:05:52,150 dogs and cats in this case. 84 00:05:52,170 --> 00:05:53,730 There is an order that makes sense. 85 00:05:53,730 --> 00:05:55,830 It should be cold medium that hot. 86 00:05:55,950 --> 00:05:59,610 You could also do hot medium cold depending on which way you want it to think about it. 87 00:05:59,640 --> 00:06:07,050 This case will say cold comes first then medium than hot so I'll go ahead and add in the level's argument 88 00:06:07,560 --> 00:06:14,130 and the levels argument is going to take in a vector and you just pasan a vector of the order you want 89 00:06:15,030 --> 00:06:20,040 needs to be ordered it's going to say cold medium and then hot 90 00:06:23,000 --> 00:06:23,320 right. 91 00:06:23,370 --> 00:06:26,160 So we've assigned that factored version of that. 92 00:06:26,370 --> 00:06:32,420 Dr. chanels go ahead and see an idea of what this looks like by calling fact that temp. 93 00:06:32,460 --> 00:06:38,010 So notice the levels actually have an order to them and you can see it by the comparison operator in 94 00:06:38,010 --> 00:06:42,570 between the terms so cold is less than medium and medium is less than hot. 95 00:06:42,960 --> 00:06:48,750 OK so that's pretty useful and it's going to be really useful when we end up using the summary function 96 00:06:49,230 --> 00:06:53,540 to call a summary on that object. 97 00:06:54,660 --> 00:07:00,720 So I'll go ahead and tell us cold medium and hot and the various counts of how many times those appear 98 00:07:01,290 --> 00:07:08,550 and we can show you a summary of the original temps as well to show the difference between the factored 99 00:07:09,030 --> 00:07:13,920 version of that vector versus the summary of the original. 100 00:07:13,920 --> 00:07:14,610 All right. 101 00:07:14,610 --> 00:07:16,860 So that's it for this lecture. 102 00:07:16,890 --> 00:07:22,320 Don't stress out too much about understanding everything you've gone over here during this lecture. 103 00:07:22,350 --> 00:07:26,160 We're not going to touch back on this until quite a while from now. 104 00:07:26,310 --> 00:07:31,900 I just wanted you to be aware of it since it fits in well with the ideas of vectors and matrices. 105 00:07:31,920 --> 00:07:39,210 So the only thing you should be aware of are the use of the factor function as a possibility for you 106 00:07:39,210 --> 00:07:39,930 to use. 107 00:07:40,080 --> 00:07:47,890 And then the difference between a nominal category set versus an ordinal category set. 108 00:07:47,910 --> 00:07:53,520 OK so the nominal you have those cats and dogs are the order ordinal as the name implies there is an 109 00:07:53,550 --> 00:07:54,480 order to it. 110 00:07:54,480 --> 00:08:00,750 And you can specify that order using the factor function by saying order equals true and then specifying 111 00:08:00,780 --> 00:08:08,230 the levels as a vector and you just pasand the vector of the terms in indexed order that you want. 112 00:08:08,510 --> 00:08:10,920 OK that's it for this lecture. 113 00:08:10,920 --> 00:08:11,960 Thanks everyone. 114 00:08:12,000 --> 00:08:13,550 I'll see at the next lecture. 11673