1 00:00:00,560 --> 00:00:05,320 My name is Ashish Mahabal and we'll be continuing with R and looking at ggplot. 2 00:00:05,320 --> 00:00:09,870 So, much of this material is taken from Headley's book by the name ggplot2. 3 00:00:09,870 --> 00:00:13,580 We start with the anatomy of a plot. 4 00:00:13,580 --> 00:00:18,160 Last time we saw a little bit of what grammar of graphics is, so this shows 5 00:00:18,160 --> 00:00:23,640 graphically how the elements in the grammar of graphics can be brought about. 6 00:00:23,640 --> 00:00:28,010 So you see at the top, that you have mapping of variables to aesthetics. 7 00:00:28,010 --> 00:00:33,230 So overall, what happens, is that you have a data set, a data frame as we saw, and 8 00:00:33,230 --> 00:00:37,350 a certain set of mappings, one or more, and then there are multiple layers. 9 00:00:38,580 --> 00:00:42,610 And for each of those layer, you again have some data which is 10 00:00:42,610 --> 00:00:46,000 very likely a subset of the original data that you have. 11 00:00:46,000 --> 00:00:50,830 Some mapping, and associated with that, a geom and a stat and 12 00:00:50,830 --> 00:00:55,210 a position where that particular layer or subset should go. 13 00:00:56,400 --> 00:01:01,059 Then, for the overall plot, you have got one scale for 14 00:01:01,059 --> 00:01:06,020 mapping, and then you've got a coordinate system and a faceting specification. 15 00:01:06,020 --> 00:01:09,190 So the coordinate system would take care of whether you have log scale or 16 00:01:09,190 --> 00:01:10,960 linear scale and so on. 17 00:01:10,960 --> 00:01:14,555 And faceting specification is where you will know what kind of 18 00:01:14,555 --> 00:01:18,780 categorical variables are being used to separate the data sets into. 19 00:01:18,780 --> 00:01:21,550 Either sublayers or subplots and so on. 20 00:01:21,550 --> 00:01:27,280 So this overall picture is very good for allowing you to debug your plot. 21 00:01:27,280 --> 00:01:31,150 So if you are trying to build a plot using layers and you're not getting what you 22 00:01:31,150 --> 00:01:34,254 want, then you can ask yourself whether you have the correct geom, 23 00:01:34,254 --> 00:01:35,960 whether you have got the correct stat, and 24 00:01:35,960 --> 00:01:39,360 how they should be interacting with each other, where should the position be, 25 00:01:39,360 --> 00:01:42,580 whether that is correct, is one of your coordinate system wrong and son on. 26 00:01:42,580 --> 00:01:45,180 So by just keeping this figure in front of you, 27 00:01:45,180 --> 00:01:49,770 many times it's fairly easy to debug what is it that you want to look at. 28 00:01:49,770 --> 00:01:52,570 Se there it is again a summary of what a plot is made of. 29 00:01:52,570 --> 00:01:56,740 Again, there is a data set, a set of mappings besides geom set and so on. 30 00:01:56,740 --> 00:01:57,980 Right? 31 00:01:57,980 --> 00:02:01,190 And so, example of scale can be this. 32 00:02:01,190 --> 00:02:04,510 It can be a continuous variable or a discrete variable. 33 00:02:04,510 --> 00:02:09,940 Remember that, that discrete variables are the ones that we use for doing 34 00:02:09,940 --> 00:02:15,260 categorization and faceting, but even the continuous variables, you can convert 35 00:02:15,260 --> 00:02:21,440 them using the as.factor function into a discrete variable and use that as well. 36 00:02:21,440 --> 00:02:26,010 So, what is it that actually happens when you make an object? 37 00:02:26,010 --> 00:02:31,090 We saw that you can assign your ggplot command to an object called b and 38 00:02:31,090 --> 00:02:34,480 not have anything rendered immediately onto the screen. 39 00:02:34,480 --> 00:02:37,340 So what is it that you can do with such an object? 40 00:02:37,340 --> 00:02:41,930 If b is the object that you have made, then you can print it. 41 00:02:41,930 --> 00:02:44,075 And this is something that have been scored automatically if 42 00:02:44,075 --> 00:02:45,460 you're not inside a loop. 43 00:02:45,460 --> 00:02:49,690 Say if you make this object and simply say P on the command line. 44 00:02:49,690 --> 00:02:52,170 It's going to do the printing of P. 45 00:02:53,410 --> 00:02:55,070 Then you can save that object. 46 00:02:55,070 --> 00:02:57,890 So if you save it will write out an image for you. 47 00:02:57,890 --> 00:02:59,810 So, rather than going to the screen, 48 00:02:59,810 --> 00:03:03,350 this is what will allow you to write it out to the disk. 49 00:03:03,350 --> 00:03:07,400 The other interesting thing is summary, it is as before or loaded and with 50 00:03:07,400 --> 00:03:12,390 this object b if you simply, say summary with empty parentheses, then you're going 51 00:03:12,390 --> 00:03:17,170 to see what is hidden in the object and we'll do that to all the objects. 52 00:03:17,170 --> 00:03:21,430 Similarly, save instead of ggsave is going to save the object itself, 53 00:03:21,430 --> 00:03:22,900 rather than as a plot. 54 00:03:22,900 --> 00:03:28,060 The object as a R object will be saved to the disc which you can reload later on. 55 00:03:28,060 --> 00:03:31,430 So let's look at the summary of the P object that we have defined at 56 00:03:31,430 --> 00:03:32,120 the top here. 57 00:03:32,120 --> 00:03:34,970 Where we are again taking the crts data set and 58 00:03:34,970 --> 00:03:38,350 simply plotting amplitude versus standard deviation there and 59 00:03:38,350 --> 00:03:42,800 factoring it into different colors by the object variable. 60 00:03:42,800 --> 00:03:47,020 So you see that though we have used only amplitude and standard deviation in 61 00:03:47,020 --> 00:03:51,030 the plot the object knows about all the variables that are in the data frame. 62 00:03:51,030 --> 00:03:53,040 So all of them are listed here. 63 00:03:53,040 --> 00:03:55,240 It also tells you what is the size of the data frame. 64 00:03:55,240 --> 00:03:57,300 There are 1619 rows in total with 26 variables. 65 00:03:57,300 --> 00:04:01,530 And then it tells you what the mapping is for the X's or 66 00:04:01,530 --> 00:04:04,849 the Y's and what facetting had been used here. 67 00:04:05,870 --> 00:04:10,532 Now in addition to p, if you were to do a geom_point, like let's do p 68 00:04:10,532 --> 00:04:15,210 + geom_point and look for the summary of that, you'll additionally see what, 69 00:04:15,210 --> 00:04:19,340 the stat_identity is, what the geom_point is, and what the position_identity is. 70 00:04:19,340 --> 00:04:23,580 In this case we have not specified that explicitly, so we get mostly NULLs, and 71 00:04:23,580 --> 00:04:24,320 FALSE, and so on. 72 00:04:24,320 --> 00:04:26,970 But all that information in principle is there. 73 00:04:26,970 --> 00:04:30,910 So that is another useful thing about the laring aspect. 74 00:04:30,910 --> 00:04:34,410 You can keep on piling up various layers into an object, and 75 00:04:34,410 --> 00:04:38,170 then those layers become available to you in various different ways. 76 00:04:39,580 --> 00:04:43,850 So, you can say save p and file name to save the data file, and 77 00:04:43,850 --> 00:04:47,070 then you can reload it again using the load command. 78 00:04:47,070 --> 00:04:49,500 And then after that you can add various layers to that, 79 00:04:49,500 --> 00:04:52,340 so, that's a very convenient thing to do. 80 00:04:52,340 --> 00:04:56,440 This is another way example of adding layers. 81 00:04:56,440 --> 00:05:00,670 So we use here ggplot, and we are using a slightly different set here. 82 00:05:00,670 --> 00:05:06,040 The set is called diamonds, and it is available with the ggplot2 library. 83 00:05:06,040 --> 00:05:10,330 So if you have started ggplot2 library the data diamonds will allow you 84 00:05:10,330 --> 00:05:12,000 to access these commands. 85 00:05:12,000 --> 00:05:17,420 So here we take that set of diamonds and use the aesthetics of number of carats 86 00:05:17,420 --> 00:05:24,100 in that and price, and then use a cut for the colors into the plot there. 87 00:05:24,100 --> 00:05:29,630 And so now if we do geom of you can add that as a layer. 88 00:05:29,630 --> 00:05:32,060 Notice here that we have used slightly different syntax. 89 00:05:32,060 --> 00:05:33,530 Rather than saying gm_point, 90 00:05:33,530 --> 00:05:37,770 we say that we want a layer and then geom equal to point. 91 00:05:37,770 --> 00:05:42,570 Now, this is where you can start combining different stats in geoms effectively. 92 00:05:42,570 --> 00:05:47,670 So, next what we are saying is that we can simply plot diamonds with aesthetics of 93 00:05:47,670 --> 00:05:51,150 x equal to carrot, so it's simply going to use that particular variable. 94 00:05:52,430 --> 00:05:55,770 But then to that p, to that object that we just defined, 95 00:05:55,770 --> 00:06:00,140 we can add a layer with a separate geom and a separate stat. 96 00:06:00,140 --> 00:06:03,520 So when you plot a histogram, what is it that is actually happening? 97 00:06:03,520 --> 00:06:07,070 You bend the data, so that is what you're, statistics is. 98 00:06:07,070 --> 00:06:09,010 And what do you actually plot? 99 00:06:09,010 --> 00:06:12,130 You plot bars of it, so that is what your geom is, 100 00:06:12,130 --> 00:06:16,850 and so this pairing is automatically done from, for histogram. 101 00:06:16,850 --> 00:06:20,330 And then when you say simply p, that is what is going to appear for 102 00:06:20,330 --> 00:06:22,100 you on the screen. 103 00:06:22,100 --> 00:06:25,583 So here is an example of a histogram as it comes out of ggplot2. 104 00:06:26,900 --> 00:06:29,614 In this case we have use a binwidth of 0.3. 105 00:06:31,780 --> 00:06:35,080 We can of course use quite interesting and 106 00:06:35,080 --> 00:06:39,030 sometimes useless stats, or geoms for a particular data set here. 107 00:06:39,030 --> 00:06:42,140 I am simply listing a few different geoms like bar, 108 00:06:42,140 --> 00:06:45,140 boxplot, contour, line, point, step, text. 109 00:06:45,140 --> 00:06:47,390 There are a total 29 of these, and just for 110 00:06:47,390 --> 00:06:52,600 fun here, I have got, used the geom of polygons on this particular data set. 111 00:06:52,600 --> 00:06:57,480 This is the same dataset that we used in the last lecture, the CIT six class set. 112 00:06:57,480 --> 00:06:59,520 And then you can see that it doesn't make sense. 113 00:06:59,520 --> 00:07:03,690 But the shape is roughly the same as what we had got when we had plotted x versus y. 114 00:07:05,400 --> 00:07:08,060 Similarly, there are 15 different statistics. 115 00:07:08,060 --> 00:07:09,290 I have listed a few here. 116 00:07:09,290 --> 00:07:11,390 The book and various websites will have more. 117 00:07:11,390 --> 00:07:16,650 We will also have it on the website of the course, so we can have contours there, 118 00:07:16,650 --> 00:07:20,420 we can have density2d, and that is what I've used here. 119 00:07:20,420 --> 00:07:25,930 I've used density2d, and I've said that let use bandwidth that is very small, 120 00:07:25,930 --> 00:07:31,840 only 0.01, and that is why you see that towards the lower parts of xy area, 121 00:07:31,840 --> 00:07:34,280 you see that there are very dense contours. 122 00:07:34,280 --> 00:07:38,220 As you go out, there is less density in, you can see that in the contours there. 123 00:07:40,080 --> 00:07:42,440 So some additional points. 124 00:07:42,440 --> 00:07:46,170 You can have the ggplot with geom_point as we saw. 125 00:07:46,170 --> 00:07:47,910 When you want to add aesthetics, 126 00:07:47,910 --> 00:07:53,050 you can give that as as arguments to the geom_point 2. 127 00:07:53,050 --> 00:07:55,450 So, if you say geon.color equals to red, 128 00:07:55,450 --> 00:07:58,420 then the points are going to be plotted as red. 129 00:07:58,420 --> 00:08:02,686 Similarly, in the layers, you can have various smoothing things. 130 00:08:02,686 --> 00:08:05,465 So here are shown that for the point method so 131 00:08:05,465 --> 00:08:10,170 that you can provide method equal to lm as your layer, and 132 00:08:10,170 --> 00:08:14,610 then smoothing can also be done addition, in addition to that. 133 00:08:15,640 --> 00:08:18,190 Similarly, you can do, scales and axes. 134 00:08:18,190 --> 00:08:21,080 We have not seen an example of that, but that's homework. 135 00:08:21,080 --> 00:08:22,380 You can take a look at that. 136 00:08:22,380 --> 00:08:24,760 We are adding that as additional layers. 137 00:08:24,760 --> 00:08:28,370 So when we say geom point, it's going to take whatever the default scale is. 138 00:08:28,370 --> 00:08:31,515 But in addition, you say that oh, I want log scale for x, 139 00:08:31,515 --> 00:08:34,980 that's sure enough add that as a layer and you get that. 140 00:08:34,980 --> 00:08:38,680 And this particular example, we have added both x and y as log scale. 141 00:08:40,280 --> 00:08:42,530 Then similarly, you can add plot options, so 142 00:08:42,530 --> 00:08:44,780 you don't have to be stuck with the defaults. 143 00:08:44,780 --> 00:08:50,223 The defaults are reasonable, but normally for making plots into publications, so 144 00:08:50,223 --> 00:08:54,640 you want the access and labels et cetera, to be much bigger, so you can specify all 145 00:08:54,640 --> 00:08:58,470 of that as a additional layer simply by using the command called opts. 146 00:08:58,470 --> 00:09:00,500 Here, I have specified for instance, 147 00:09:00,500 --> 00:09:03,520 what the title should be and what the aspect ratio should be. 148 00:09:03,520 --> 00:09:04,990 And you can add various other commands with that. 149 00:09:06,900 --> 00:09:12,890 So, the homework with respect to this is, take the dataset that we have provided and 150 00:09:12,890 --> 00:09:17,410 we saw examples using only two variables, amplitude and standard deviation. 151 00:09:17,410 --> 00:09:20,210 But there are several other interesting variables. 152 00:09:20,210 --> 00:09:24,900 Try to see if you can use them and separate them in many different ways. 153 00:09:24,900 --> 00:09:28,570 This, what we are covered here is fairly basic graphics. 154 00:09:28,570 --> 00:09:33,600 We were not trying to visualize in any very different or 155 00:09:33,600 --> 00:09:35,860 specific way how to go about things. 156 00:09:35,860 --> 00:09:38,020 We were looking at only what is available. 157 00:09:38,020 --> 00:09:41,080 There will be additional modules in the course where you'll be looking at 158 00:09:41,080 --> 00:09:44,620 visualization, how you should be plotting the data, et cetera. 159 00:09:45,930 --> 00:09:46,430 Thank you.