WEBVTT 1 00:00:01.081 --> 00:00:02.312 Hi guys, and welcome back. 2 00:00:02.312 --> 00:00:04.200 In this video, we're going to learn how to draw 3 00:00:04.200 --> 00:00:07.010 a stacked bar chart using matplotlib. 4 00:00:07.010 --> 00:00:08.150 Over the last couple of videos, 5 00:00:08.150 --> 00:00:10.950 we've explored how to integrate matplotlib 6 00:00:10.950 --> 00:00:13.960 into a application that uses a database. 7 00:00:13.960 --> 00:00:16.150 Now, because we've already learned how to do that, 8 00:00:16.150 --> 00:00:18.000 I'm going to start moving away from that, 9 00:00:18.000 --> 00:00:19.850 so that I don't have to write the same code 10 00:00:19.850 --> 00:00:22.370 and queries over and over that we have been writing 11 00:00:22.370 --> 00:00:23.980 through the entire course and instead, 12 00:00:23.980 --> 00:00:27.510 we're going to focus simply on the matplotlib aspect itself. 13 00:00:27.510 --> 00:00:30.560 So for this lecture, I've given you the data, 14 00:00:30.560 --> 00:00:33.050 and you can copy that and paste it into a file 15 00:00:33.050 --> 00:00:34.180 in your project. 16 00:00:34.180 --> 00:00:37.600 I'm naming that file data.py in my case, 17 00:00:37.600 --> 00:00:39.940 and it contains some information that we'll need 18 00:00:39.940 --> 00:00:41.730 for this bar chart. 19 00:00:41.730 --> 00:00:43.310 Let's get started. 20 00:00:43.310 --> 00:00:45.510 What we've got in data.py, which is linked 21 00:00:45.510 --> 00:00:47.490 in the resources section of this lecture 22 00:00:47.490 --> 00:00:49.250 is this set of polls. 23 00:00:49.250 --> 00:00:52.930 Each poll has a name, the number of men who voted 24 00:00:52.930 --> 00:00:55.610 in this poll and the number of women who voted in this poll. 25 00:00:55.610 --> 00:00:58.130 Normally when you're collecting vote information, 26 00:00:58.130 --> 00:01:00.980 you're gonna collect more than just male or female, 27 00:01:00.980 --> 00:01:03.120 you may collect more information about gender, 28 00:01:03.120 --> 00:01:06.140 you may collect information about age, and where they live, 29 00:01:06.140 --> 00:01:08.460 et cetera, all that demographic data. 30 00:01:08.460 --> 00:01:10.720 But for this case, since we only want to show you 31 00:01:10.720 --> 00:01:13.950 how to plot a stacked bar chart, this is enough information. 32 00:01:13.950 --> 00:01:16.400 So here we've got the polls list 33 00:01:16.400 --> 00:01:19.560 and that is what we're going to import in app.py. 34 00:01:19.560 --> 00:01:23.740 So I'm gonna go ahead and do import matplotlib.pyplot as plt 35 00:01:23.740 --> 00:01:26.270 and from data import polls. 36 00:01:26.270 --> 00:01:30.300 And now we can go ahead and draw our stacked bar chart. 37 00:01:30.300 --> 00:01:32.710 We're gonna start off with just a simple bar chart, 38 00:01:32.710 --> 00:01:35.810 a normal bar chart, so we'll do poll_titles, 39 00:01:35.810 --> 00:01:38.320 we need this for the axis labels, 40 00:01:38.320 --> 00:01:43.080 poll zero for poll in polls. 41 00:01:43.080 --> 00:01:47.340 And we need the poll men, which is poll one 42 00:01:47.340 --> 00:01:49.720 for poll in polls. 43 00:01:49.720 --> 00:01:51.490 Now that we've got this, we can use that 44 00:01:51.490 --> 00:01:53.240 to create our bar chart. 45 00:01:53.240 --> 00:01:57.020 We also need the x coordinates of each bar, 46 00:01:57.020 --> 00:01:59.800 which I'm gonna create as poll x coordinates 47 00:01:59.800 --> 00:02:02.960 and this is range of len of polls. 48 00:02:02.960 --> 00:02:05.190 Now we can go ahead and use matplotlib 49 00:02:05.190 --> 00:02:08.560 to create the bar chart with figure equal plt.figure. 50 00:02:08.560 --> 00:02:12.270 The fixed size, I'm gonna set to six, six. 51 00:02:12.270 --> 00:02:14.420 Then we'll create our set of axes, 52 00:02:14.420 --> 00:02:17.090 which is figure.add sub plot. 53 00:02:17.090 --> 00:02:19.070 And then we can do axes.bar, 54 00:02:19.070 --> 00:02:23.140 and in here we'll put the x coordinates, the poll men, 55 00:02:23.140 --> 00:02:26.980 and as tick label, we'll use the poll titles. 56 00:02:26.980 --> 00:02:29.670 Then at the end, we can do plt.show. 57 00:02:29.670 --> 00:02:32.950 Nothing new here, we've essentially grouped everything 58 00:02:32.950 --> 00:02:34.830 that we know about matplotlib so far, 59 00:02:34.830 --> 00:02:36.610 and we've used it to create our bar chart. 60 00:02:36.610 --> 00:02:38.930 So let me go ahead and run this file 61 00:02:38.930 --> 00:02:42.720 and you can see that we get our bar chart here. 62 00:02:42.720 --> 00:02:45.540 Of course, we have the problem where the poll titles 63 00:02:45.540 --> 00:02:48.470 are not rotated and so forth, but we already know 64 00:02:48.470 --> 00:02:50.230 how to rotate them. 65 00:02:50.230 --> 00:02:51.600 So instead of focusing on that, 66 00:02:51.600 --> 00:02:53.940 we're going to stack our bar charts. 67 00:02:53.940 --> 00:02:56.700 The first thing we need is to get the women voters, 68 00:02:56.700 --> 00:03:01.600 so we'll do poll_women poll two for poll in polls 69 00:03:01.600 --> 00:03:05.290 and then inside the same set of axes, 70 00:03:05.290 --> 00:03:07.390 we're gonna draw another bar chart. 71 00:03:07.390 --> 00:03:11.210 So we'll say axes.bar, and again, x coordinates, 72 00:03:11.210 --> 00:03:14.940 the poll women, the tick label is going to be the same, 73 00:03:14.940 --> 00:03:18.960 which is poll titles, and we're gonna add one more thing, 74 00:03:18.960 --> 00:03:21.300 but before we add that thing, let me just run this 75 00:03:21.300 --> 00:03:22.670 as it is right now. 76 00:03:22.670 --> 00:03:25.260 And you can see that what we get here 77 00:03:25.260 --> 00:03:28.730 is something that looks like a stacked bar chart, right? 78 00:03:28.730 --> 00:03:31.270 But these three here in the middle are not stacked, 79 00:03:31.270 --> 00:03:32.103 so why not? 80 00:03:32.103 --> 00:03:35.970 What we're drawing here is we're drawing the men first 81 00:03:35.970 --> 00:03:38.210 and then the women in front if you will, 82 00:03:38.210 --> 00:03:41.010 if we are the camera, looking at this chart, 83 00:03:41.010 --> 00:03:43.290 the women are in front and they are hiding 84 00:03:43.290 --> 00:03:44.570 some of the men votes. 85 00:03:44.570 --> 00:03:48.270 So the total of this chart or what it goes up to 86 00:03:48.270 --> 00:03:52.340 is 60 because that is the higher of the two votes. 87 00:03:52.340 --> 00:03:54.510 However, for example, in the second poll, 88 00:03:54.510 --> 00:03:57.510 we only see orange because the women voted more. 89 00:03:57.510 --> 00:04:00.530 The 15 is down here, hidden behind it, 90 00:04:00.530 --> 00:04:02.290 so this is not a stacked bar chart, 91 00:04:02.290 --> 00:04:05.447 this is a pretty useless bar chart, if anything, 92 00:04:05.447 --> 00:04:06.380 so we do have to make a small change 93 00:04:06.380 --> 00:04:08.760 to put one on top of the other instead of in front 94 00:04:08.760 --> 00:04:10.298 of the other. 95 00:04:10.298 --> 00:04:11.270 And what we'll do is very straightforward, 96 00:04:11.270 --> 00:04:14.475 bottom equal poll men. 97 00:04:14.475 --> 00:04:18.620 So notice that poll_men is the values of the men votes, 98 00:04:18.620 --> 00:04:20.983 and what we're seeing in here when we're plotting 99 00:04:20.983 --> 00:04:23.722 the women votes, is that they should appear 100 00:04:23.722 --> 00:04:26.015 on top of the men votes. 101 00:04:26.015 --> 00:04:28.110 Essentially, what matplotlib is doing behind the scenes, 102 00:04:28.110 --> 00:04:32.300 is it's adding to the women votes, the men votes, 103 00:04:32.300 --> 00:04:34.300 and then it appears on top. 104 00:04:34.300 --> 00:04:37.453 So if we run this, you'll see that that is what we get. 105 00:04:38.357 --> 00:04:40.870 Now, let's learn how to adjust the x tick labels. 106 00:04:40.870 --> 00:04:43.341 We've learned how to do that earlier on, 107 00:04:43.341 --> 00:04:45.180 but we are going to do something slightly different, 108 00:04:45.180 --> 00:04:47.450 because you can see that both our plots 109 00:04:47.450 --> 00:04:50.948 have a tick_label property. 110 00:04:50.948 --> 00:04:54.297 And what happens when we do this is that the latest one 111 00:04:54.297 --> 00:04:56.000 is the one that shows up in the axes 112 00:04:56.000 --> 00:04:58.631 and in this case, they're both the same 113 00:04:58.631 --> 00:04:59.720 so we don't really mind which one shows up 114 00:04:59.720 --> 00:05:02.476 but in other situations we might. 115 00:05:02.476 --> 00:05:05.000 So it's a bit confusing to pass the tick label 116 00:05:05.000 --> 00:05:07.839 into both bar polls. 117 00:05:07.839 --> 00:05:09.079 And also, it's a bit wasteful 118 00:05:09.079 --> 00:05:10.634 because we have that duplication. 119 00:05:10.634 --> 00:05:12.210 So instead, what we can do is we can get rid of this 120 00:05:12.210 --> 00:05:17.120 and that and instead use plt.xticks to pass in 121 00:05:17.965 --> 00:05:20.391 the x coordinates, the labels as well 122 00:05:20.391 --> 00:05:22.421 as any other configuration property. 123 00:05:22.421 --> 00:05:24.662 So we'll pass in here poll x coordinates, 124 00:05:24.662 --> 00:05:28.400 we need this to map each label to one bar, 125 00:05:28.400 --> 00:05:29.990 then we'll pass in the titles. 126 00:05:29.990 --> 00:05:32.666 It's important that these two are the same length 127 00:05:32.666 --> 00:05:34.611 because the first label will be mapped 128 00:05:34.611 --> 00:05:36.725 to the first coordinate and the second label 129 00:05:36.725 --> 00:05:37.730 to the second coordinate and so on. 130 00:05:37.730 --> 00:05:39.050 Then something else that we can do 131 00:05:39.050 --> 00:05:40.760 is pass in the rotation 132 00:05:40.760 --> 00:05:44.269 and the horizontal alignment as well. 133 00:05:44.269 --> 00:05:45.910 Now let me run this again and you'll see that we get 134 00:05:45.910 --> 00:05:47.770 our bar chart like that. 135 00:05:47.770 --> 00:05:49.890 You might see that the bar chart is cut off 136 00:05:49.890 --> 00:05:51.860 at the bottom though, because that's where the subplot's 137 00:05:51.860 --> 00:05:53.120 adjust comes into play. 138 00:05:53.120 --> 00:05:56.890 Let's give our x axis ticks a bit more room with that. 139 00:05:56.890 --> 00:06:00.140 Here we can do figure.subplots adjust 140 00:06:00.140 --> 00:06:02.890 bottom equals 0.35. 141 00:06:02.890 --> 00:06:04.920 And now when we run this, we've got a bit of room 142 00:06:04.920 --> 00:06:07.570 for our labels there to show up. 143 00:06:07.570 --> 00:06:09.810 Alright, that was a bit of new information, 144 00:06:09.810 --> 00:06:12.370 with drawing the bar chart and also some review 145 00:06:12.370 --> 00:06:13.400 of old information. 146 00:06:13.400 --> 00:06:16.270 Hopefully everything's coming together nicely for you guys. 147 00:06:16.270 --> 00:06:17.470 Thank you for joining me in this video. 148 00:06:17.470 --> 00:06:20.223 Thanks for watching, and I'll see you in the next one.