WEBVTT 00:00.550 --> 00:04.090 In this video we learn how to remove complete rows from a data frame. 00:04.090 --> 00:11.800 But first of all we import pandas and we import our Summer Olympics data set and reset the athlete column 00:11.800 --> 00:21.210 s so the index and we have a look at the first 10 rows and then let's assume that we want to remove 00:21.300 --> 00:24.570 all instances of Alfred Hi us. 00:24.570 --> 00:30.090 So we have here the estimate column is our index and we have two appearances of three titles and we 00:30.090 --> 00:38.440 want to remove these appearances from our data frame and same with columns we can use and the drop method 00:38.530 --> 00:44.560 and we can now pass to the index parameter the index labels that we want to drop. 00:44.560 --> 00:50.980 So in this case free tiles and we can try to do this. 00:51.740 --> 00:58.580 So here we can see we have dropped off for tiles and let's say again our data frame and it's still there 00:58.580 --> 01:06.350 because we need to set up the in-place perimeter to true but first of all let us decided that we not 01:06.350 --> 01:13.470 only want to remove all instances of life for the tiles but also we want to remove auto Hirschman and 01:13.530 --> 01:19.990 same with columns we can pass a list of labels to the parameter index. 01:20.220 --> 01:25.590 And if you want to change our complete data frame we have to set the in-place parameter to Drew. 01:25.620 --> 01:27.600 So let's do this. 01:28.330 --> 01:31.120 And now let's have a look at our data frame. 01:33.030 --> 01:37.090 So we can see we have to remove TR for tires and auto Ashman 01:40.880 --> 01:43.090 and the same with removing columns. 01:43.100 --> 01:45.880 We have fear for the drop method to alternatives. 01:45.890 --> 01:54.020 So either we can pass our index labels to the parameter index or we can posit to the pyramid does labels 01:54.590 --> 01:56.330 and then we define the axis. 01:56.330 --> 02:05.430 So in this case and the axis is zero or index we can use both either 0 or index in quotation marks and 02:05.430 --> 02:11.190 again if we want to change our data frame we have to set the in-place perimeter to true. 02:11.190 --> 02:12.200 So let's do this 02:15.460 --> 02:23.740 and let's check again if we have now removed the rovers Dimitrios give us yes. 02:23.820 --> 02:25.330 So this works. 02:25.590 --> 02:31.590 So selecting or dropping rows by Row Labels is not something that you will use very often. 02:31.590 --> 02:37.470 So it's more likely to select or drop rows them that meet on do not meet certain conditions and some 02:37.470 --> 02:38.790 columns. 02:38.790 --> 02:43.930 So selecting and removing rows by billion indexing as they are actually the more frequent case. 02:44.310 --> 02:46.080 And let's assume we laughed. 02:46.110 --> 02:49.590 The Summer Olympic Games in 1996 in Atlanta. 02:49.620 --> 02:55.640 And so what we only want to have rows of medals from the 1996 Olympic Games. 02:55.650 --> 03:03.870 And we can do this by selecting all rows that the year column is 1996 and then we can override our summer 03:03.870 --> 03:13.760 data frame and let's see what we get here let's check our new summer data frame and here we have only 03:13.760 --> 03:19.180 rows where we have the value 1996 in the year column. 03:19.190 --> 03:25.160 All right so now let's re import our whole summer Olympics data frame and set the athlete columnist 03:25.160 --> 03:32.440 to the index so now again we have our original data frame. 03:32.490 --> 03:37.500 Now let's assume that we do not want to select a rose that meets certain conditions but we want to drop 03:37.500 --> 03:39.840 rose and that meets certain conditions. 03:39.840 --> 03:46.950 So let's assume we are not interested in the Olympic Games 1996 and we do not like aquatics so we want 03:46.950 --> 03:48.560 to remove all rows there. 03:48.570 --> 03:55.620 The years either 1996 or the sports Aquatics and therefore we create two brilliant serious and then 03:55.620 --> 04:02.970 the first one we check whether the year is 1996 and we assign the variable mask one and then the second 04:02.970 --> 04:10.110 condition we check whether the sport is Aquatics and we assign the variable mask to and then in a first 04:10.110 --> 04:10.410 step. 04:10.410 --> 04:15.840 I filter all rows that meet either condition one or either condition two. 04:15.960 --> 04:23.490 So by having a summer top log and then mask one or mask two I would have all rows where at least one 04:23.490 --> 04:30.700 condition here is Matt however I want to select all rows are neither of the two conditions are met and 04:30.700 --> 04:34.840 therefore I have to use the two the symbol here. 04:34.840 --> 04:41.740 And by doing so I get a data frame with rows where the year is neither 1996 nor the sports columnist 04:41.770 --> 04:49.450 Aquatics and we can override our original data frame and set our new data frame okay. 04:49.460 --> 04:52.790 First of all we have defined our to a mosque one and mosque two. 04:52.820 --> 04:54.000 Okay. 04:54.290 --> 04:56.630 Now this would work and let's have a look here. 04:59.230 --> 05:04.210 So now at a first glance we cannot really see if they're still ninety nine in six and our year column 05:04.240 --> 05:10.830 or aquatics in our spot column and that's actually four ways to do this. 05:10.840 --> 05:16.930 So first of all we check the year column and we create a boolean serious and check whether in the year 05:16.930 --> 05:22.020 column desk nine 1996 and then we make a value account so let's check this. 05:22.600 --> 05:27.490 And you know year column we have actually twenty five thousand three hundred ninety eight forces and 05:27.480 --> 05:36.610 no true so we have removed all instances of ninety nine and six in the year column and we could also 05:36.610 --> 05:43.660 do we could transform our year column into a an umpire Ray this summer to adopt values and check whether 05:43.660 --> 05:51.470 ninety nine in six is in the name pi array and this is false. 05:51.540 --> 05:59.480 All right so we have removed all rows with the ninety nine in six in the year column and let's check 05:59.510 --> 06:04.250 if he's still at the Aquatics and the sports column and we can do this by checking whether in the spot 06:04.250 --> 06:11.050 column there's still Aquatics and we are using the Utzon method so let's try this out here 06:15.070 --> 06:21.670 and we get a boolean serious and now we can check if any of the values in our brilliant serious is a 06:21.670 --> 06:32.220 true and we can do this with the any method and this gives us a false or that's actually no true in 06:32.230 --> 06:37.480 our boolean serious so we have removed all instances of Aquatics 06:40.210 --> 06:45.550 and we can actually do the same and creating the boolean serious by checking whether aquatics is in 06:45.550 --> 06:51.690 the sports column and again we can check whether at least the one true is here in our pavilion serious 06:51.850 --> 06:55.840 with the any method and again we get a false. 06:55.840 --> 07:00.850 So there's no row with aquatics in the sports column. 07:00.990 --> 07:05.650 All right so now this is it for the time being and I hope to see you in the next session by.