WEBVTT 00:01.070 --> 00:06.330 Hi everybody I saw in the last video we learned how to select a single rose with the look operator. 00:06.350 --> 00:12.440 And in this video we will further enhance our skills and learn how to slice the rows and columns service 00:12.440 --> 00:14.240 through the local operator. 00:14.240 --> 00:22.340 But before we start we import pandas and we import our battle data set and set to the indexed column 00:23.020 --> 00:23.840 to athletes. 00:23.870 --> 00:29.320 So athlete is our index column and let's have a look here at the first five rows. 00:29.330 --> 00:37.530 So we have every athlete columnist index and then the last video we learned that was the operator log. 00:37.560 --> 00:41.370 We can make a label based indexing so we can pass a label. 00:41.370 --> 00:48.080 So for example Michael Phelps within this BAP records so let's see what you got here 00:52.020 --> 00:59.960 and we got here a data set to this all rose that contains the Michael Phelps on the index so if the 00:59.960 --> 01:04.800 label yeah of course many times so we get all occurrences of the label here 01:08.190 --> 01:10.940 and same as with the i log operator. 01:10.960 --> 01:15.570 So the first position here within the brackets is for the index. 01:15.570 --> 01:20.130 And the second position here after the commerce for the column. 01:20.130 --> 01:25.310 So what's the first position you index your slice in the index and with the second position. 01:25.500 --> 01:32.410 You slice our index in the column and our flag sample we can select only the column metal 01:37.130 --> 01:45.020 and we get a panda serious with them all occurrences of Michael Phelps s label of our series here and 01:45.030 --> 01:53.380 as the values uh we get all medals off the rose son the name of the panda series will see a medal. 01:53.380 --> 02:03.190 All right so let's copy it here and paste it in our next line and with the local operator we can of 02:03.190 --> 02:06.720 course also slice the for many column. 02:06.720 --> 02:15.490 So for example we want to have a medal and we want to have the event as bracket here is missing. 02:17.230 --> 02:18.790 All right. 02:18.790 --> 02:21.520 So now we get all medals of Michael Phelps. 02:21.560 --> 02:25.300 Um and we get them the kind of medal and the kind of event. 02:25.930 --> 02:26.680 So here. 02:26.680 --> 02:35.280 Michael Phelps won the gold medal for 100 meter butterfly and we can also select for more than one label 02:35.280 --> 02:35.510 here. 02:35.520 --> 02:42.090 So let's copy it here and let's say we're not only want to get all medals and the events of Michael 02:42.090 --> 02:45.480 Phelps but also we are interested in let's say Carl Lewis. 02:46.320 --> 02:51.870 So we open here a list and here we have Michael Phelps. 02:51.870 --> 02:54.770 And uh let's uh try here. 02:54.780 --> 02:55.860 Carl Lewis 03:01.020 --> 03:02.760 and let's close our list 03:05.600 --> 03:08.360 so here we are slicing for Michael Phelps and Carl Lewis. 03:08.390 --> 03:11.050 And we want to know the medals and the event. 03:13.310 --> 03:15.800 And here we get here we have Michael Phelps. 03:15.800 --> 03:19.340 And here we have all occurrences of Carl Lewis. 03:19.820 --> 03:23.750 So he won a gold in 100 meter 200 meter long jump. 03:23.780 --> 03:24.350 And so on 03:29.060 --> 03:34.030 all right so now let's go back and let's have a look at the first 10 rows here. 03:34.030 --> 03:35.150 So we had method 03:39.110 --> 03:44.330 and with the local operator we not only can select four particular rows and columns we can also make 03:44.330 --> 03:45.860 slices. 03:45.890 --> 03:50.280 So for example what if we want to slice from the very beginning. 03:50.420 --> 03:54.580 So from the first draw and two spirit on comes up is slow. 03:54.590 --> 03:55.250 Let's try it. 03:55.280 --> 04:02.480 So we take your summer dot log and then within the back brackets we say OK we want to all rose from 04:02.480 --> 04:04.420 the beginning until. 04:05.330 --> 04:07.220 And let's copy it here. 04:11.840 --> 04:15.280 So let's make clear quotation marks and Let's paste. 04:15.470 --> 04:23.400 So let's see what we got here so we got all the rows from the beginning of flat tires until spirit on 04:23.590 --> 04:24.600 just. 04:24.660 --> 04:32.400 And the very last column here is including So typically with the position based indexing or slicing 04:32.400 --> 04:33.690 like with the i log. 04:33.690 --> 04:40.670 Operator we learned that the starting point is including and the ending point is excluding but now with 04:40.670 --> 04:42.680 them the label base indexing. 04:42.840 --> 04:49.100 So both inside including so we have the starting point including and also the ending point yes be redone 04:49.190 --> 04:54.270 up this is also including and spirit on jazz up. 04:54.270 --> 04:57.690 Is this actually a new unique value in our index. 04:57.690 --> 05:03.990 So there's only one appearance of spirit on Charles up is and let's try to do the same with the flight 05:03.990 --> 05:08.820 sample Michael Phelps we have we know there are many occurrences of Michael Phelps 05:19.970 --> 05:26.780 so now here we can see that we get an error message and it says that cannot get the right Sly's bound 05:26.780 --> 05:28.180 for non unique labels. 05:28.190 --> 05:36.300 Michael Phelps so label based the slicing with the labels that are not unique and does not work here. 05:36.320 --> 05:39.550 All right so and we can also try the following here. 05:39.560 --> 05:40.810 So let's copy it 05:44.720 --> 05:55.190 and um before we saw here that Alfred high US has two occurrences so the first uh row and the first 05:55.190 --> 05:58.330 second third fourth fifth sixth seventh row. 05:58.390 --> 06:05.180 So now if we want to slice and the whole dataset beginning from high retires until the very end. 06:05.210 --> 06:09.390 We can try this so let's copy here the label 06:17.820 --> 06:23.340 so we want to have all rows beginning from us at high us until the very end. 06:23.730 --> 06:30.150 And also here we get an error message cannot get left slice bond fund on unique label Alfred high as 06:31.320 --> 06:35.620 so label based sizing only works with a unique label. 06:35.970 --> 06:42.400 All right so let's move on and let's have a look at uh the twenty first. 06:42.390 --> 06:43.410 The rows here 06:47.450 --> 06:54.080 and what we could do we could actually slice from and uh beginning point to an ending point so let's 06:54.710 --> 07:03.270 take summer download and then we could say we want to slice from Dimitrios to leave us here 07:09.670 --> 07:13.730 until for example Arthur Blake 07:21.030 --> 07:25.020 and I can tell you that both labels are unique and our index. 07:25.410 --> 07:26.580 So this should work. 07:29.270 --> 07:35.100 And then we can see we get a data frame so a slice of our original data frame beginning with Demetrius 07:35.160 --> 07:43.900 Rivas and ending this as a plague and both ending points including and we already know from position 07:43.900 --> 07:50.230 based indexing that by default the starting point is including but the ending points excluding so this 07:50.230 --> 07:58.160 is seer the friend with a label based indexing and we can also do a label based the slicing for the 07:58.160 --> 08:01.580 columns so let's copy it here. 08:05.770 --> 08:12.260 And let's say we want to slice for Demetrius three of us but only for Demetrius three of us and for 08:12.260 --> 08:19.610 the columns we want the columns uh from a year until support including so we could say from the column 08:19.610 --> 08:22.640 year until the columns support 08:27.960 --> 08:34.860 and here we can see that we get for the rows of the label Demetrius three of all columns beginning from 08:34.860 --> 08:36.000 year until spot. 08:36.000 --> 08:42.540 So also here the starting point and the ending point are including all right. 08:42.540 --> 08:45.660 So last but not least let's try another one here. 08:49.520 --> 08:54.280 And uh let's say we want to slice for Michael Phelps again 09:00.140 --> 09:04.730 and we want to slice additionally for let's say Donald Duck 09:10.770 --> 09:13.410 so I do think that there is no label. 09:13.530 --> 09:14.160 Donald Duck. 09:14.160 --> 09:18.630 So there was no athlete that ever won a medal was the named Donald Duck. 09:18.630 --> 09:26.720 But we can check and we can see so first of all we have here all occurrences of Michael Phelps. 09:26.760 --> 09:30.570 And then here at the very end in our new data frame we have them. 09:30.680 --> 09:34.280 They wrote on attack and in our column CMBS and they end badly. 09:34.290 --> 09:40.170 So we have missing values because in our original data frame there is no robust simply labeled on attack. 09:40.170 --> 09:47.820 But still we get here a result but of course with the missing values and piles and drops here a warning 09:49.240 --> 09:56.620 and says that if we pass labels that are actually not in our data frame in the future python will raise 09:56.620 --> 10:00.500 an arrow here and we should use them the re index method. 10:01.240 --> 10:06.910 So but for the time being now we can ignore this because I'm still at that works here. 10:07.000 --> 10:12.250 So we have Donald Duck here with any end values and all the values of Michael Phelps and we know that 10:12.300 --> 10:13.980 penthouse is still under development. 10:13.990 --> 10:18.200 So this could change in the future and therefore we have the future warning. 10:18.210 --> 10:18.670 All right. 10:18.670 --> 10:27.160 So let's copy it here and we can also try to slice four columns which are not in our original data frame 10:27.190 --> 10:28.180 so let's say. 10:28.990 --> 10:29.410 OK. 10:29.410 --> 10:31.110 MICHAEL PHELPS That's OK. 10:34.460 --> 10:37.280 And we want to slice for the columns 10:39.540 --> 10:47.230 here and we assume that now our data frame that's in the column H with the age of each athlete. 10:47.580 --> 10:50.300 So we know that there is no column age but we can try. 10:51.510 --> 10:58.040 And also here we get all occurrences of Michael Phelps here we get the year column and we get the H 10:58.050 --> 11:03.240 column which we do not have from our it's no data frame but here in our new data frame we have the H 11:03.240 --> 11:04.670 column with the missing value. 11:04.690 --> 11:09.830 So it still works and we get to the same future warning that still this works. 11:09.840 --> 11:14.910 So passing labels and that are actually not in the original data frame but in the future which should 11:14.940 --> 11:18.300 use in the re index method. 11:18.490 --> 11:18.840 All right. 11:18.850 --> 11:24.470 So this is it for this session and then next session we will have a quick summary of all that we learned 11:24.480 --> 11:25.630 in the last video. 11:25.640 --> 11:27.310 So to see it there by.