WEBVTT 00:00.970 --> 00:07.180 Hi everybody is so in this video we learn how to select rows with the index operator loke the so-called 00:07.210 --> 00:13.420 label based indexing and we will see in a minute but the differences to position based indexing with 00:13.540 --> 00:14.370 the eye log. 00:14.370 --> 00:16.420 OPERATOR So what. 00:16.420 --> 00:23.200 First of all we import pandas as always and we import our Summer Olympic Games data frame and we have 00:23.200 --> 00:29.680 a look at the first five rows and before we can select so let's assume we want to select the third row 00:29.680 --> 00:36.610 at the next position to this Dimitrios device so we can take summer thought I'd like. 00:37.510 --> 00:40.700 And then the next position too. 00:41.170 --> 00:45.870 And here we get a panda serious with the athlete the Dimitrios three of us. 00:45.930 --> 00:52.810 Won bronze in the 100 meter freestyle for sailors and the method read CSP provides many options and 00:52.810 --> 00:57.270 parameters to customize our imparting of CSP files. 00:57.280 --> 01:02.860 So on by default to the row index is a range indexed starting from zero but they are we can change this 01:02.950 --> 01:09.640 and we can determine yet each of our column and our data frame as the index column or the index. 01:09.640 --> 01:17.530 So that is the parameter index column and that we can pass a column of our data frame and determine 01:17.800 --> 01:20.040 the column s of the index of our data frame. 01:20.050 --> 01:22.420 So let's try for example athlete 01:27.230 --> 01:28.640 and let's have a check here. 01:28.640 --> 01:36.920 So now we can see that the athlete column is actually now our index off our data frame and the eye log 01:36.950 --> 01:39.380 operator still works also with a different index. 01:39.380 --> 01:42.280 So not only with the range index. 01:42.500 --> 01:48.560 So what the island operator does it simply selects in the row at the index position to for example and 01:48.680 --> 01:56.440 this is irrespective of which column we selected here as the index so let's check here again and we 01:56.440 --> 02:00.240 still get here the rovers the Dimitrios three of us. 02:00.340 --> 02:05.410 So that's no difference for the island operator if you change here the index labels. 02:05.410 --> 02:11.080 However there are some minor differences so the athlete label itself is not yet on the left side as 02:11.190 --> 02:14.920 the index label off our pandas series which we created here. 02:15.010 --> 02:17.890 And actually the whole name of our pandas seriousness. 02:17.920 --> 02:20.920 Um yeah actually DiMaggio's three of us. 02:21.070 --> 02:21.480 All right. 02:21.490 --> 02:29.740 So now let's head over to the loke operator and now we still want to select that row of Dimitrios three 02:29.740 --> 02:36.280 of us and what we have to do here we have to pass that then this back brackets the index label in this 02:36.280 --> 02:41.840 case Dimitrios three of us in quotation marks 02:46.160 --> 02:52.630 and again we get a panda serious with them the row of Dimitrios three US was so years cities Park discipline 02:52.640 --> 03:00.200 country gender event and medal as labels on the left side and the values on the right side and the whole 03:00.200 --> 03:03.170 pandas serious is called Dimitrios to leave us. 03:03.260 --> 03:09.560 All right so let's search for a very famous sports stars and we could search for Michael Phelps for 03:09.560 --> 03:10.190 example. 03:11.570 --> 03:13.700 So we have to pass in quotation marks. 03:13.730 --> 03:18.450 The name here and capital letters so helps comma. 03:18.590 --> 03:22.640 And then the first name where only the first characters and capital letters. 03:22.640 --> 03:23.660 So Michael 03:27.060 --> 03:28.230 now yes a typo. 03:28.390 --> 03:31.780 So it's my car now. 03:34.090 --> 03:35.050 No. 03:35.240 --> 03:42.320 And instead of getting only one row as a pen serious we get to hear a complete pen that stayed off frame 03:42.420 --> 03:45.100 because my good friend says one how many medals. 03:45.110 --> 03:48.220 And he appeared in the index many times. 03:48.290 --> 03:55.190 So if you select the index label Michael Phelps listed the loke operator then we get all appearances 03:55.200 --> 04:00.780 of the index label Michael Phelps and now what. 04:00.780 --> 04:07.500 If you want to select only the very first medal then we can do this by slicing the very first throw 04:07.500 --> 04:15.540 here off our new data frame by applying the I luck operator and selecting the very fast draw at an exposition 04:15.560 --> 04:16.330 zero. 04:16.380 --> 04:23.970 So I would expect that we get the gold medal from 2004 for the 100 meter butterfly event 04:26.740 --> 04:27.730 and is this correct. 04:27.730 --> 04:33.450 So now we have a pen a serious and this is actually the pitfall of the log operator. 04:33.510 --> 04:35.940 So if you have duplicates in our index. 04:35.940 --> 04:41.700 So Michael Phelps has many appearances then logo returns as a data frame. 04:41.700 --> 04:47.710 And if you have only one appearance of the labor for which we want to slice our data frame for example 04:47.720 --> 04:51.060 here Dimitrios through us we get a pan that serious. 04:51.060 --> 04:57.510 So far the same operation we get as a return different data types and that's always a bit difficult 04:57.600 --> 05:01.260 because of the functionality is different. 05:01.280 --> 05:01.660 All right. 05:01.670 --> 05:06.140 So now we are finished with this session and in the next session we will have a look at the slicing 05:06.140 --> 05:07.910 rows and columns. 05:08.050 --> 05:09.120 The low operate. 05:09.160 --> 05:10.610 So hope to see you there by.