1 00:00:00,810 --> 00:00:08,360 All right, you know how to slice lists 0 to 1 2 00:00:08,400 --> 00:00:13,880 will give you the first item of the list, and 0 to 2 as will give you the first and the second item of the list and 3 00:00:13,890 --> 00:00:16,800 so on. With Numpy arrays 4 00:00:16,800 --> 00:00:23,640 you do basically the same thing except with Numpy arrays sometimes you have two or more dimentionals 5 00:00:23,940 --> 00:00:25,560 or three. 6 00:00:25,560 --> 00:00:31,760 So let me go ahead and I'll create, I'll start with a two dimensional array. 7 00:00:32,190 --> 00:00:35,430 So I'll get that from this variable. 8 00:00:35,430 --> 00:00:38,950 So this is the Numpy array we'll be working with. 9 00:00:39,150 --> 00:00:47,540 And yet if you want to extract this number here, this here, this one and this. 10 00:00:47,670 --> 00:00:49,840 So these four numbers. 11 00:00:50,040 --> 00:00:56,040 Well you'd first need to set the index of the rows that you want to slice. 12 00:00:56,190 --> 00:01:03,340 So that would be zero to two which gives you the first two rows. 13 00:01:03,480 --> 00:01:08,660 Actually we can try that and you'll see that you get the first two rows. 14 00:01:08,670 --> 00:01:16,020 Now if you want only this portion here, so these two numbers, and these two you would want to pass 15 00:01:17,430 --> 00:01:31,910 a comma there and then the index for the columns which would be you know 0 here, 1, so 2, 3. And yeah, 2 to 4 maybe. Yeah, that should do it. 16 00:01:32,350 --> 00:01:35,100 And that's it. 17 00:01:35,220 --> 00:01:37,610 That means you have this indexing system. 18 00:01:37,610 --> 00:01:45,110 So you start from 0 for rows, then the second row is 1 and 2 and so on. For columns the same thing. 19 00:01:45,380 --> 00:01:50,120 0 1 2 3 and 4. 20 00:01:50,540 --> 00:01:55,310 And actually if you like you can see the shape of your Numpy array. 21 00:01:55,550 --> 00:02:01,010 So it's 3 by 5 which means 0 to 2, and 0 to 4. 22 00:02:01,160 --> 00:02:01,780 Great. 23 00:02:02,090 --> 00:02:06,410 And of course you can use the convention of list indexing. 24 00:02:06,650 --> 00:02:11,950 So you can pass zero and everything after that or just like that. 25 00:02:12,020 --> 00:02:17,780 So you get all the rows and only the columns from two to three. 26 00:02:17,800 --> 00:02:24,240 Similarly you can get only one value if you like. 27 00:02:24,460 --> 00:02:34,520 Not particularly this one because three is out of the bounds for axis zero, so axis zero is a horizontal axis which 28 00:02:34,550 --> 00:02:37,950 says that there is no row with index 3. 29 00:02:38,320 --> 00:02:41,370 So we said it's 0 1 2. 30 00:02:41,470 --> 00:02:47,500 So let's pass 2 there and you get 182 which is the very last 31 00:02:47,500 --> 00:02:47,850 item 32 00:02:47,870 --> 00:02:49,340 of the Numpy array. 33 00:02:49,830 --> 00:02:51,360 Yeah that's about indexing. 34 00:02:51,430 --> 00:02:54,170 How about iterating through a Numpy array? 35 00:02:54,850 --> 00:02:56,820 Well there are two ways to do that. 36 00:02:56,830 --> 00:03:09,730 The first way is to say, let's say for i in your Numpy array print i, and what that will do is it will print 37 00:03:09,730 --> 00:03:14,410 out the rows of your Numpy arrays. 38 00:03:14,470 --> 00:03:18,900 So what this i axis is in your Numpy arrays is the rows. 39 00:03:18,970 --> 00:03:24,670 So each the ration it gets the first row and it prints that out in the first iteration, then the second 40 00:03:24,670 --> 00:03:27,770 one, and then the third one. 41 00:03:27,840 --> 00:03:35,530 And if you want to iterate through columns you'd want to access the transposed version of your Numpy array 42 00:03:36,460 --> 00:03:38,260 and that's how you do it. 43 00:03:38,710 --> 00:03:49,040 And also if you want to iterate value by value, you'd say for i in im_g.flat 44 00:03:51,830 --> 00:03:55,670 print i. 45 00:03:55,800 --> 00:04:00,190 So this property what it does is it allows you to access the values 46 00:04:00,190 --> 00:04:06,130 of your Numpy array one by one. And that's about indexing, slicing, and iterating. 47 00:04:06,230 --> 00:04:07,600 I'll see you in the next lecture!