WEBVTT 00:00.510 --> 00:04.220 In this video we learn how to select columns from a data frame. 00:04.230 --> 00:12.180 But before we have to import Parnassus PD as always then we import our Titanic data frame and assign 00:12.180 --> 00:17.210 the variable Titanic and we will have a look here at the first five throws. 00:17.220 --> 00:18.710 We already know this. 00:18.960 --> 00:22.820 And then let's assume that we are only interested in the H column. 00:22.830 --> 00:29.850 So we want to analyze them the age of all passengers that are aboard the Titanic and we can do this 00:29.850 --> 00:32.760 by riding Titanic here. 00:32.760 --> 00:34.740 So that's all a data frame. 00:34.740 --> 00:41.130 And then we open the scrap racket and then we pass the label off our column that we want to select. 00:41.130 --> 00:49.330 So in this example it's some H and we have to posit in quotation marks. 00:49.390 --> 00:56.200 So now here we select the H column let's run the sail here and you can see we got all values within 00:56.200 --> 01:03.640 the H column at the same order as in our data frame and still on the left here we have the index labels 01:03.700 --> 01:05.130 off our data frame. 01:05.290 --> 01:14.350 So at index position 0 or label 0 we have 22 as an H and journal label the one we have from a peasant 01:14.350 --> 01:14.990 service. 01:15.070 --> 01:17.410 Thirty eight years old and so on 01:21.980 --> 01:28.730 and we can also check that the data type here off our column so let's copy and paste and then let's 01:28.730 --> 01:30.140 check in the data type 01:34.680 --> 01:41.930 and here we get a panda serious so the pen that serious is also that data Tidewater class within the 01:41.930 --> 01:49.130 panda's library and it's one dimensional label array so you can see here we have as an atlas or name 01:49.220 --> 01:52.790 one dimensional name pi array one sequence of values. 01:52.790 --> 02:00.440 So we have one dimension here and so here on the left we have a label and as I said before so the serious 02:00.440 --> 02:07.740 is also a class all data type which has specific attributes and methods which we will learn later and 02:07.740 --> 02:10.980 most of the methods and attributes are quite similar to a data frame. 02:10.980 --> 02:17.760 So there are some methods some attributes that we can only apply with a pen that serious and the other 02:17.760 --> 02:23.790 way around there are some methods some attributes that can only apply to Panda's data frames. 02:23.790 --> 02:30.780 So now what if we want to select a column but we want to still have it as a data frame so let's paste 02:30.780 --> 02:32.280 again here our column. 02:32.280 --> 02:39.330 And if you want to have a data frame instead of a serious we can pass inside this back brackets our 02:39.570 --> 02:44.470 H column with another aspect bracket here. 02:44.700 --> 02:50.850 So we have doubled our bracket and the fear around this year and we can also see from the layout that 02:50.970 --> 03:05.150 this could be a data frame and that's copy and paste and let's check in the type here. 03:05.280 --> 03:10.700 So in this manner we are still a data frame by having two spat brackets here and depending on what you 03:10.700 --> 03:16.360 want to do with them the column you can either selected as a pen that serious or as a data frame. 03:17.540 --> 03:18.010 All right. 03:18.020 --> 03:18.710 So now what. 03:18.710 --> 03:23.000 If you want to select the two columns. 03:23.000 --> 03:33.590 So let's assume you want to have the edge column and the sex column somebody tried to do this by passing 03:33.590 --> 03:39.790 age and sex in within the span bracket and you would get an error message. 03:41.090 --> 03:48.470 So something is wrong here and the solution is that if you want to select many columns we have to pass 03:48.560 --> 03:52.600 them as a list into our spam brackets. 03:52.970 --> 04:01.250 So we have to pass them the list with age and sex to our staff brackets and let's see what we got here. 04:02.970 --> 04:08.850 Now we got here a data frame with them the age column and the sex column and still on the left side 04:09.390 --> 04:17.870 we have the row labels and we can also rearrange here the order off the column so let's copy here and 04:17.870 --> 04:24.260 paste again can rearrange them the order of the columns the fear change and the sequence here and our 04:24.260 --> 04:25.070 list. 04:25.190 --> 04:31.080 So let's say we want to hear the sex column before the age column. 04:31.760 --> 04:32.420 We can do this. 04:32.450 --> 04:37.930 So now we have them first to the sex column and the column compared to here 04:42.080 --> 04:46.100 and of course we can also select multiple columns. 04:46.360 --> 04:52.000 So let's copy and paste and we can also include here within our list. 04:52.100 --> 05:01.880 The fact column now we get a pen this data frame with three columns six H and fair and therefore the 05:01.880 --> 05:11.750 same order and the original data frame so let's check again the original data frame the first rows and 05:11.750 --> 05:12.720 then we can see it. 05:12.740 --> 05:21.050 So the first row we have a male which is 22 years old and paid a fare of seven point two five pound 05:21.080 --> 05:24.780 or dollar and the second row here. 05:24.820 --> 05:31.690 We have a female thirty eight years old and the female paid seventy one point two dollars or pounds 05:33.250 --> 05:36.620 so you can see here the order of our rose and then our slice. 05:36.630 --> 05:44.010 Yes exactly the same as in the original data frame and that's a second way to select one column. 05:44.040 --> 05:52.680 So either we do this by passing our column within the span brackets or we can just use the pandas attribute 05:52.710 --> 05:55.950 dot and then the name or label of the column. 05:55.950 --> 06:03.360 So for example h here so we have h as the column label and automatically in our data frame this and 06:03.360 --> 06:08.350 the attribute top eight so let's see what we get here. 06:09.190 --> 06:09.320 Yeah. 06:09.350 --> 06:13.040 And we actually got the same pattern as serious with our H column. 06:13.040 --> 06:18.680 So here we have all the values of our each column and also here on the left side um the labels 06:22.300 --> 06:29.200 and last but not least we can check if both alternatives are actually the same 06:32.070 --> 06:39.120 so we can check if I'm here I will see areas that we created within the attribute H equals 06:41.690 --> 06:49.850 and then we can pass on the second alternative so titanic and passing H and we get a true here. 06:49.880 --> 06:56.810 So the method equal checks whether the pen does serious Titanic that HCR actually the same s and the 06:56.810 --> 07:02.420 pen as serious and that we created here was pap records and it's actually a matter of taste if you select 07:02.420 --> 07:09.420 a column with them the attribute or the spell brackets and later we will learn that there are some minor 07:09.420 --> 07:17.700 differences however for just the selecting a column and both works the same way and in my opinion the 07:17.910 --> 07:20.630 attributes and sometimes a bit more comfortable. 07:20.640 --> 07:30.550 So if fewer type a titanic thought and if you want to select the let's say the impact column we press 07:30.580 --> 07:38.110 M and then we benefit from the tab completion function so now if you press the the tab key then Python 07:38.140 --> 07:44.560 automatically checks which attributes are available and uh see some that there's some the impact column 07:45.400 --> 07:49.900 and then we have here the impact column as a panacea is. 07:50.680 --> 07:54.420 So this is quite comfortable actually. 07:54.460 --> 07:54.760 All right. 07:54.760 --> 07:59.460 So now we have finished the session and in the next session we learn how to select growth. 07:59.470 --> 08:00.660 Hope to see you there by.