1 00:00:00,000 --> 00:00:08,530 DAVID JORDAN: Hello, and welcome back to recitation. 2 00:00:08,530 --> 00:00:11,990 The problem I'd like to work with you now is simply to 3 00:00:11,990 --> 00:00:14,030 compute some partial derivatives using the 4 00:00:14,030 --> 00:00:16,490 definitions we learned today in lecture. 5 00:00:16,490 --> 00:00:20,150 So first we're going to compute the partial derivative 6 00:00:20,150 --> 00:00:23,790 in the x-direction, this function xy squared 7 00:00:23,790 --> 00:00:25,250 plus x squared y. 8 00:00:25,250 --> 00:00:26,860 Then we're going to compute its derivative in the 9 00:00:26,860 --> 00:00:30,270 y-direction, and then finally we're going to evaluate the 10 00:00:30,270 --> 00:00:32,760 partial derivative in the x-direction at a particular 11 00:00:32,760 --> 00:00:36,080 point 1, 2 for the first problem, and in the second 12 00:00:36,080 --> 00:00:38,260 problem, we're going to compute second partial 13 00:00:38,260 --> 00:00:38,760 derivatives. 14 00:00:38,760 --> 00:00:42,400 Now these we just compute by taking the derivative of the 15 00:00:42,400 --> 00:00:45,020 derivative, just as we do in one-variable calculus. 16 00:00:45,020 --> 00:00:48,330 So why don't you work on these, pause the tape, and 17 00:00:48,330 --> 00:00:49,940 I'll check back in a moment, and we'll 18 00:00:49,940 --> 00:00:51,190 see how I solve these. 19 00:00:51,190 --> 00:00:59,300 20 00:00:59,300 --> 00:01:00,150 OK, welcome back. 21 00:01:00,150 --> 00:01:01,460 Let's get started. 22 00:01:01,460 --> 00:01:06,750 So we have x squared y-- 23 00:01:06,750 --> 00:01:11,620 excuse me, xy squared plus x squared y. 24 00:01:11,620 --> 00:01:13,950 That's our f. 25 00:01:13,950 --> 00:01:17,120 So when we take the partial derivative in the 26 00:01:17,120 --> 00:01:18,410 x-direction-- 27 00:01:18,410 --> 00:01:21,820 remember, this just means that we treat y as if it were a 28 00:01:21,820 --> 00:01:24,620 constant, and we just take an ordinary derivative in the 29 00:01:24,620 --> 00:01:27,440 x-direction as we would do in one-variable calculus. 30 00:01:27,440 --> 00:01:30,940 So the derivative of this in the x-direction is just y 31 00:01:30,940 --> 00:01:35,430 squared, because we only differentiate the x here. 32 00:01:35,430 --> 00:01:41,130 Similarly here, the derivative of x squared is 2x, and y just 33 00:01:41,130 --> 00:01:43,100 comes along for the ride as if it were a constant. 34 00:01:43,100 --> 00:01:48,520 35 00:01:48,520 --> 00:01:50,670 For the partial derivative in the y-direction, we do the 36 00:01:50,670 --> 00:01:54,680 same thing, except now, x is a constant, and we're taking an 37 00:01:54,680 --> 00:01:56,950 ordinary derivative in the y-direction. 38 00:01:56,950 --> 00:02:03,610 So we have 2xy plus x squared. 39 00:02:03,610 --> 00:02:06,670 40 00:02:06,670 --> 00:02:10,520 And then the final thing that we need to do is we want to 41 00:02:10,520 --> 00:02:18,390 evaluate partial f, partial x at the point 1, 2. 42 00:02:18,390 --> 00:02:22,200 And so all that means is that we have to plug in x equals 1 43 00:02:22,200 --> 00:02:26,360 and y equals 2 into our previous computation, and so 44 00:02:26,360 --> 00:02:32,170 we get 2 squared plus 2 times 1 times 2. 45 00:02:32,170 --> 00:02:37,240 So altogether, we get 8. 46 00:02:37,240 --> 00:02:40,850 So that's computing partial derivatives. 47 00:02:40,850 --> 00:02:43,530 Now let's move on and compute the second partial 48 00:02:43,530 --> 00:02:45,720 derivatives. 49 00:02:45,720 --> 00:02:50,610 So, for instance, we want to compute the second partial 50 00:02:50,610 --> 00:02:56,780 derivative both times in the x-direction. 51 00:02:56,780 --> 00:02:59,790 So all this means is that when we took the first partial, we 52 00:02:59,790 --> 00:03:01,780 got a function of x and y, and now we just 53 00:03:01,780 --> 00:03:03,370 need to take its partial. 54 00:03:03,370 --> 00:03:06,240 So we just need to take the derivative of this again in 55 00:03:06,240 --> 00:03:07,400 the x-direction. 56 00:03:07,400 --> 00:03:09,900 So now, the derivative of y squared-- 57 00:03:09,900 --> 00:03:12,640 be careful-- the derivative of y squared in the x-direction 58 00:03:12,640 --> 00:03:16,660 is just zero, because y is a constant relative to x. 59 00:03:16,660 --> 00:03:20,322 And so, then altogether, we just get 2y. 60 00:03:20,322 --> 00:03:24,010 When we take the derivative to this x, we just get one. 61 00:03:24,010 --> 00:03:29,550 So that's our second partial derivative in the x-direction. 62 00:03:29,550 --> 00:03:35,505 And now you can also take mixed partials. 63 00:03:35,505 --> 00:03:39,750 64 00:03:39,750 --> 00:03:44,590 So here, we take a derivative of f. 65 00:03:44,590 --> 00:03:46,880 First we take the derivative in the y-direction and then we 66 00:03:46,880 --> 00:03:50,100 take a derivative of that in the x-direction. 67 00:03:50,100 --> 00:03:52,500 So we can look at our derivative here, partial f, 68 00:03:52,500 --> 00:03:55,140 partial y, and we need to take its partial in the 69 00:03:55,140 --> 00:03:56,470 x-direction. 70 00:03:56,470 --> 00:04:03,370 And so we get 2y plus 2x. 71 00:04:03,370 --> 00:04:07,770 72 00:04:07,770 --> 00:04:10,250 Now let's see what happens if we switch the order here and 73 00:04:10,250 --> 00:04:19,580 we take, instead, the partial derivative in 74 00:04:19,580 --> 00:04:20,860 the opposite order. 75 00:04:20,860 --> 00:04:24,330 So now let's go back to our partial derivative of f in the 76 00:04:24,330 --> 00:04:28,400 x-direction and let's take its derivative now in the 77 00:04:28,400 --> 00:04:30,330 y-direction. 78 00:04:30,330 --> 00:04:35,090 So the first term there, y squared, gives us a 2y and the 79 00:04:35,090 --> 00:04:38,540 second term gives us a 2x. 80 00:04:38,540 --> 00:04:43,680 I want to just note that these are equal. 81 00:04:43,680 --> 00:04:47,470 In fact, the mixed partial derivatives, whether you take 82 00:04:47,470 --> 00:04:53,980 them in the xy order or the yx order for the sorts of 83 00:04:53,980 --> 00:04:56,640 functions that we're going to be considering in this class, 84 00:04:56,640 --> 00:05:00,090 for instance, all polynomial functions and all 85 00:05:00,090 --> 00:05:03,510 differentiable functions of several variables, these mixed 86 00:05:03,510 --> 00:05:05,180 partials are going to be equal. 87 00:05:05,180 --> 00:05:08,150 In your textbook, there are some examples of sort of 88 00:05:08,150 --> 00:05:12,410 pathological functions where these are not equal, but 89 00:05:12,410 --> 00:05:14,440 certainly for any polynomial functions, these are always 90 00:05:14,440 --> 00:05:16,290 going to be equal. 91 00:05:16,290 --> 00:05:18,600 And I think I'll leave it at that. 92 00:05:18,600 --> 00:05:18,785