WEBVTT 1 00:00:00.200 --> 00:00:01.680 Hi guys, and welcome back. 2 00:00:01.680 --> 00:00:03.400 In this video, we're going to learn how to add 3 00:00:03.400 --> 00:00:06.140 multiple axes to one figure, so that we can 4 00:00:06.140 --> 00:00:09.540 end up exporting images that have multiple graphs in them. 5 00:00:09.540 --> 00:00:10.914 Let's get started. 6 00:00:10.914 --> 00:00:14.220 Here we've got some of the code from the last video. 7 00:00:14.220 --> 00:00:16.390 We've got our figure that we're creating, 8 00:00:16.390 --> 00:00:20.350 and then we're adding a subplot, which is the axes, 9 00:00:20.350 --> 00:00:23.400 then in there, we're plotting our line chart. 10 00:00:23.400 --> 00:00:25.820 What we can also do though is we can rename 11 00:00:25.820 --> 00:00:29.720 this to, let's say, AX1 for axis one and axis two, 12 00:00:29.720 --> 00:00:31.610 and we're gonna do add subplot again. 13 00:00:31.610 --> 00:00:34.400 Now we do have two subplots in our figure, 14 00:00:34.400 --> 00:00:36.660 and we can grab that and plot something again. 15 00:00:36.660 --> 00:00:39.120 So I'm gonna plot, in this case, a slightly 16 00:00:39.120 --> 00:00:41.050 different set of numbers, one, two, three, 17 00:00:41.050 --> 00:00:43.570 four in the X axis, but five, seven, 18 00:00:43.570 --> 00:00:46.020 eleven, and seventeen in the Y axis. 19 00:00:46.020 --> 00:00:48.000 Now, if we run this, you might expect, 20 00:00:48.000 --> 00:00:49.110 since you've already seen it, 21 00:00:49.110 --> 00:00:51.320 and I've already mentioned, that you'll have 22 00:00:51.320 --> 00:00:53.720 two sets of axes in your figure, 23 00:00:53.720 --> 00:00:56.770 but that's actually not what's going to happen at all. 24 00:00:56.770 --> 00:01:00.800 Here we have one pair of axes, and two plots in it. 25 00:01:00.800 --> 00:01:01.700 So what's happened? 26 00:01:01.700 --> 00:01:05.030 Well by default, when we do figure.add subplot, 27 00:01:05.030 --> 00:01:07.590 matplotlib is going to add the subplot 28 00:01:07.590 --> 00:01:10.990 in the top left hand corner of our figure. 29 00:01:10.990 --> 00:01:13.860 Figures are essentially divided into cells. 30 00:01:13.860 --> 00:01:17.100 The top left hand corner is cell number one, 31 00:01:17.100 --> 00:01:19.290 the next one to its right is cell number two, 32 00:01:19.290 --> 00:01:20.920 the next one to its right is cell number three, 33 00:01:20.920 --> 00:01:24.060 and so on, and it goes left to right, top to bottom, 34 00:01:24.060 --> 00:01:26.640 like the diagram on the screen is currently showing. 35 00:01:26.640 --> 00:01:29.290 But here we've not told matplotlib to create 36 00:01:29.290 --> 00:01:31.210 any number of rows and columns. 37 00:01:31.210 --> 00:01:33.730 We've only told it to create different subplots. 38 00:01:33.730 --> 00:01:37.250 So by default, they're both the same subplot. 39 00:01:37.250 --> 00:01:39.700 If we want to specify different subplots, 40 00:01:39.700 --> 00:01:42.860 we can use the nomenclature of number of rows, 41 00:01:42.860 --> 00:01:47.190 comma, number of columns, comma, the index for this axes. 42 00:01:47.190 --> 00:01:51.220 So where this axes is going to appear within that grid. 43 00:01:51.220 --> 00:01:53.350 Here we're defining that this axis 44 00:01:53.350 --> 00:01:55.530 will be the top left hand corner one, 45 00:01:55.530 --> 00:01:58.170 in a grid of two columns. 46 00:01:58.170 --> 00:02:00.810 So this one is gonna be one, two, two, 47 00:02:00.810 --> 00:02:02.710 to be the one at the top right, 48 00:02:02.710 --> 00:02:04.722 in a grid of two columns. 49 00:02:04.722 --> 00:02:07.540 Now if we do this, we can run this again, 50 00:02:07.540 --> 00:02:09.800 and you see that we get two different graphs. 51 00:02:09.800 --> 00:02:11.330 Now the lines are also the same colour, 52 00:02:11.330 --> 00:02:13.270 because matplotlib is aware that these 53 00:02:13.270 --> 00:02:15.560 are two separate plots in two different places. 54 00:02:15.560 --> 00:02:17.200 The user is not gonna get confused 55 00:02:17.200 --> 00:02:19.414 between which one is which. 56 00:02:19.414 --> 00:02:23.990 When working with matplotlib, you'll often see this. 57 00:02:23.990 --> 00:02:26.620 One two one, one two two, just like that. 58 00:02:26.620 --> 00:02:28.120 It means exactly the same thing, 59 00:02:28.120 --> 00:02:30.210 but matplotlib allows us to pass 60 00:02:30.210 --> 00:02:32.510 in a single set of numbers, as apposed 61 00:02:32.510 --> 00:02:34.590 to three comma separated values. 62 00:02:34.590 --> 00:02:37.210 I really don't like this at all, 63 00:02:37.210 --> 00:02:39.520 and I would encourage you not to use it, 64 00:02:39.520 --> 00:02:41.000 but it is something that you're gonna see 65 00:02:41.000 --> 00:02:43.544 every now and then, when working with matplotlib code. 66 00:02:43.544 --> 00:02:46.560 And obviously it doesn't work when the number 67 00:02:46.560 --> 00:02:48.820 of cells reaches a number greater than nine, 68 00:02:48.820 --> 00:02:51.200 because then you need four digits instead of three. 69 00:02:51.200 --> 00:02:53.260 So generally I would recommend using 70 00:02:53.260 --> 00:02:55.750 the three number approach, with three 71 00:02:55.750 --> 00:02:59.660 different arguments, rather than passing in a single number. 72 00:02:59.660 --> 00:03:02.290 Here we have created a figure and two axes, 73 00:03:02.290 --> 00:03:04.810 and we can do this using this piece of code. 74 00:03:04.810 --> 00:03:06.910 However, there is a shorthand that we can use 75 00:03:06.910 --> 00:03:08.970 when working with multiple subplots. 76 00:03:08.970 --> 00:03:11.560 Instead of three different variables there, 77 00:03:11.560 --> 00:03:16.560 we can do figure, X1, X2, equal plt.subplots, 78 00:03:17.060 --> 00:03:19.020 and that is going to create one figure 79 00:03:19.020 --> 00:03:22.060 with two subplots, in this same structure. 80 00:03:22.060 --> 00:03:23.867 AX1 is gonna be the first one. 81 00:03:23.867 --> 00:03:25.500 AX2 is gonna be the second one. 82 00:03:25.500 --> 00:03:26.750 Note that I'm getting these errors here, 83 00:03:26.750 --> 00:03:28.210 because the variables are re-declared, 84 00:03:28.210 --> 00:03:30.510 but if we get rid of those, that goes away. 85 00:03:30.510 --> 00:03:32.790 If you start wanting to have multiple 86 00:03:32.790 --> 00:03:36.180 rows and columns, like for example two and two, 87 00:03:36.180 --> 00:03:38.350 then you're gonna start running into some problems, 88 00:03:38.350 --> 00:03:43.350 because AX1 and AX2 will now start meaning each row. 89 00:03:43.890 --> 00:03:47.074 So AX1 will be the first row, with two plots. 90 00:03:47.074 --> 00:03:49.880 X2 will be the second row, with two plots. 91 00:03:49.880 --> 00:03:53.613 So instead what we have to do is something like this. 92 00:03:56.280 --> 00:03:59.470 So here we have AX1 and AX2, and then in a separate 93 00:03:59.470 --> 00:04:01.290 tuple, AX3 and AX4. 94 00:04:01.290 --> 00:04:03.070 If you're not familiar with tuple, 95 00:04:03.070 --> 00:04:04.030 the structuring by the way, I've got 96 00:04:04.030 --> 00:04:05.990 a resource linked in the resources section 97 00:04:05.990 --> 00:04:07.520 of this lecture, for you to check out. 98 00:04:07.520 --> 00:04:09.830 What's happening here is that plt.subplots 99 00:04:09.830 --> 00:04:14.270 is returning a figure, and a tuple of tuples, 100 00:04:14.270 --> 00:04:15.950 and we are then de-structuring that 101 00:04:15.950 --> 00:04:18.560 tuple of tuples into four different variables. 102 00:04:18.560 --> 00:04:20.270 Again, if we go back to one two, 103 00:04:20.270 --> 00:04:23.660 and we do AX1 and AX2, subplots 104 00:04:23.660 --> 00:04:25.900 is now returning a tuple with figure 105 00:04:25.900 --> 00:04:28.100 and two axes, and we're de-structuring 106 00:04:28.100 --> 00:04:30.470 those two into their two variables. 107 00:04:30.470 --> 00:04:32.199 This works in exactly the same way. 108 00:04:32.199 --> 00:04:34.030 All right, that's everything for this video. 109 00:04:34.030 --> 00:04:35.360 Thank you guys for joining me on this one. 110 00:04:35.360 --> 00:04:36.550 I hope you've learned something, 111 00:04:36.550 --> 00:04:38.300 and you'll see you in the next one.