WEBVTT 00:01.730 --> 00:07.580 In this video we will quote some more examples love split apply combine and by doing so you want to 00:07.580 --> 00:13.230 get more insight into our data and we want to find more interesting facts and relationships. 00:13.250 --> 00:18.810 All right so now let's first of all import pandas and now we are working with the Summer Olympics data 00:18.870 --> 00:19.250 said 00:24.230 --> 00:30.080 and now we want to find the 20 most successful countries in the history of Summer Olympic Games with 00:30.140 --> 00:34.100 the highest number of medals and therefore we have to split here. 00:34.100 --> 00:40.910 Our summer data frame by the column or by the key country and by doing so we are getting about 140 40 00:41.060 --> 00:48.020 different groups our countries and for each group we want to count the amount of medals that were already 00:48.020 --> 00:50.810 awarded to athletes from the country. 00:51.410 --> 00:57.740 And therefore we are only interested in the medal column and for each group we want to count the amount 00:57.740 --> 01:05.870 of non values in the medal column and we can do this with the count method and we are starting the results 01:05.900 --> 01:10.940 in the variable medals per country so let's do this and also print medals per country. 01:12.200 --> 01:18.800 And we can see here on the left side we have the country and on the right side the total amount of medals 01:18.920 --> 01:22.550 that were awarded to athletes from the respective country. 01:22.880 --> 01:31.130 So for example Afghanistan has two medals then here Belgium has four hundred and eleven medals. 01:31.130 --> 01:35.290 And Brazil has 400 31 medals and so on. 01:35.420 --> 01:41.100 And what we actually wanted to know is actually the 20 most successful countries. 01:41.150 --> 01:46.130 So what we can do now actually we can change the year the method and just 01:50.170 --> 01:53.430 and we want to have the 20 countries with the most medals. 01:53.440 --> 01:56.230 So let's pass here and equals 20. 01:57.310 --> 02:00.590 And by running the side we are actually overriding the rebels. 02:00.610 --> 02:01.480 So let's do this 02:05.480 --> 02:08.540 A.B. After the 20 most successful countries. 02:08.540 --> 02:15.410 So no surprise that the most successful country is the USA and the second most successful country I 02:15.410 --> 02:18.340 think this is the former Soviet Union. 02:18.350 --> 02:27.050 And then we have great Britain France Germany Italy Australia Hungary Sweden other lands and so on so 02:27.050 --> 02:35.400 actually on ranked 20 is Denmark with five hundred and seven medals and we can also plot our results 02:35.410 --> 02:37.110 yet in a bar chart. 02:37.190 --> 02:40.550 So let's first of all import plot lib as always. 02:40.550 --> 02:42.320 And we use in the seaborne stated 02:45.280 --> 02:48.730 and then we are plotting here our pandas serious matters per country. 02:48.750 --> 02:57.430 And we are using a bar chart and the figure size is 14 aid and we are using a font size of 14 and we 02:57.430 --> 03:03.880 f at the X label of country the violator the number of medals and the titles the Summer Olympic Games 03:03.910 --> 03:05.200 total medals per country. 03:05.200 --> 03:06.820 So let's see what we get here 03:11.180 --> 03:17.230 and here we have the bar chart of the 20 most successful countries in some Olympic Games. 03:17.230 --> 03:24.640 So number one this year USA was over 4000 medals then we have the former Soviet Union Great Britain 03:24.640 --> 03:28.270 France and to Poland and Denmark. 03:28.270 --> 03:28.740 All right. 03:28.750 --> 03:31.780 So now let's head over to our Titanic data dataset 03:36.040 --> 03:43.090 and we can use here to describe method to get some more summary statistics on our North America columns 03:46.080 --> 03:54.890 so here we can see for example that the average fare that OP hasn't just paid was uh 30 to and we can 03:54.890 --> 03:57.890 also apply the mean method directly on the fare column. 03:57.890 --> 04:01.710 So also give you should get thirty two point two. 04:02.060 --> 04:05.560 And this is actually the average fare for all passengers. 04:05.570 --> 04:12.020 But now let's assume that we want to know the average fare for all passenger classes and we can do this 04:12.500 --> 04:16.160 by splitting our data frame into the passenger classes. 04:16.160 --> 04:22.760 So we have three passenger classes then we only select select the fare column and we want to know the 04:22.760 --> 04:26.840 average are the mean fare paid by the three passenger classes. 04:26.840 --> 04:28.070 So let's do this here. 04:29.000 --> 04:34.190 And it's actually no surprise then that the average fare for the first class is highest. 04:34.190 --> 04:42.170 So 84 than the average fare paid by a second class passengers is twenty point six and third class passengers 04:42.170 --> 04:45.800 paid on average thirteen point six. 04:45.800 --> 04:53.270 And we can also take the sum over the whole survival column so we might recall that in these last column 04:53.300 --> 04:57.510 we have one fossilized and 0 4 not survived. 04:57.860 --> 05:03.540 And by taking the sum over these survived column we should get the total amount of passengers who actually 05:03.540 --> 05:05.590 is left to disaster. 05:05.600 --> 05:07.480 So let's do this. 05:07.520 --> 05:10.940 So in total a three hundred forty two passengers survived. 05:11.570 --> 05:18.540 And by calculating the mean on the survive column we are actually calculating the survival rate. 05:18.830 --> 05:25.160 So one stands for survived 0 4 not survived and by taking the mean we are actually getting these survival 05:25.160 --> 05:25.520 rate 05:28.300 --> 05:33.410 so only thirty eight point three of our peasant system survived the disaster. 05:33.410 --> 05:37.360 And we know that there was a clear code of conduct and filling the lifeboats. 05:37.370 --> 05:39.740 So women and children first. 05:39.830 --> 05:45.920 And now we can check if they really followed the code of conduct and we can do so by calculating the 05:45.920 --> 05:51.290 survival rate for different groups of flight sample we can calculate the survival rate for them and 05:51.290 --> 05:55.910 for man for children for adults and also for passenger classes. 05:55.910 --> 06:02.900 And as a first step let's calculate the survival rate by gender so the survival rate of all few male 06:02.900 --> 06:04.760 passengers and on male passengers 06:07.580 --> 06:10.730 and we can see that there is a significant difference. 06:11.000 --> 06:18.980 So the total survival rate of all passengers was 38 percent but the survival rate for the female passengers 06:19.040 --> 06:25.250 were 74 percent and for all male passengers only 18 or 19 percent. 06:25.310 --> 06:29.060 And we can also check the survival rate by passenger class 06:33.420 --> 06:39.990 and we can see that first and second class passengers had some priority when it comes to getting a place 06:40.020 --> 06:41.700 in a lifeboat. 06:41.700 --> 06:48.840 So sixty three percent of all first class passengers survived and only 24 percent of all third class 06:48.840 --> 06:49.920 passengers survived. 06:49.950 --> 06:51.660 So this is a clear difference. 06:51.870 --> 06:56.030 And we can also split our data frame into adulthood and children. 06:56.070 --> 07:02.340 However we do not have a column with added ought to run and therefore we have to create a new column. 07:02.430 --> 07:05.060 So we create the column adult to run. 07:05.610 --> 07:13.170 And first of all we assign the value added to all rows and then we filter our data frame and for all 07:13.170 --> 07:20.400 rows we add the h just below 18 B overwrite the value in the edited to add one column by adding value 07:20.400 --> 07:21.120 would tie it. 07:21.120 --> 07:27.440 So let's do this here and let's have a look at our data frame with our new column. 07:27.450 --> 07:28.380 Add a child 07:31.110 --> 07:37.320 so we can see here for the first 5 percent just about are all older than 18 years old and therefore 07:37.350 --> 07:39.470 we have here the value added. 07:39.660 --> 07:42.870 But let's have a look at some more rows here 07:46.530 --> 07:51.840 so here the peasant at next position 7 is actually only two years old. 07:51.990 --> 07:54.920 And here we have the value child and the column. 07:54.930 --> 07:55.710 Adult child 07:58.460 --> 08:05.460 all right and let's also now analyze our adult child to the column by counting the values with the method 08:05.460 --> 08:14.060 to value counts so in total we have seven hundred seventy eight adults and one hundred thirteen children. 08:14.230 --> 08:21.570 And we can also split and group our Titanic data frame by the column add a child and we can calculated 08:21.610 --> 08:24.580 the survival rate for adults and for children. 08:24.600 --> 08:25.510 So let's do this 08:29.100 --> 08:37.410 so on average uh 36 percent of all adults survived and about 54 percent of our children survived the 08:37.410 --> 08:38.510 disaster. 08:38.790 --> 08:44.430 And we've already learned that we can use more than one key so we can use two or three or even more 08:44.430 --> 08:46.780 keys to split our data frame. 08:47.010 --> 08:51.960 And now as and the Code of Conduct was uh women and children first. 08:52.290 --> 09:00.000 We want to split our data framed by the sex and by the adult or child column and by doing so we can 09:00.000 --> 09:03.000 split our data frame into four groups. 09:03.240 --> 09:05.960 And you want to analyze this advice column. 09:05.970 --> 09:09.870 But first of all do you want to count the number of peasants us in each group. 09:09.870 --> 09:17.700 So let's do this here with the count method so here we can see we have four different groups. 09:17.920 --> 09:20.680 So we have the combination of female and adult. 09:20.680 --> 09:27.180 We have the combination female child and we have the combination male adult and a male child. 09:27.190 --> 09:33.640 So here we have actually a minority index or hierarchical index where the first index columnist and 09:33.640 --> 09:39.540 the sex column and the second the other child column and here on the right side we have the values. 09:39.610 --> 09:42.450 So the amount of passengers in each group. 09:42.580 --> 09:49.120 So we have here a panda serious with a hierarchical index and we want to calculate the survival rate 09:49.120 --> 09:58.290 for all four groups and instead of applying the count method we now can apply the mean method and what 09:58.290 --> 10:04.890 we can already do we can slot our four groups from the highest survival rate to the lowest survival 10:04.890 --> 10:10.770 rate with the sort values method and passing falls to the ascending parameter 10:14.050 --> 10:20.170 so we have a descending order starting with the group with the highest survival rate until the group 10:20.170 --> 10:21.850 has the lowest survival rate. 10:21.850 --> 10:29.590 So let's do this here and here we can see the group with the highest survival rate is actually female 10:29.590 --> 10:32.830 adults with 75 percent survival rate. 10:33.160 --> 10:39.290 Then the second highest survival rate is actually female child with sixty nine percent. 10:39.310 --> 10:47.590 Then we have a male child with the 39 percent and by far the lowest survival rate had actually adult 10:47.620 --> 10:52.470 male passengers with only a sixteen point five percent. 10:52.480 --> 10:59.200 So again here we have a panda serious with an hierarchical index and we can also store yeah our pandas 10:59.200 --> 11:02.440 serious and the variable of women and children first. 11:02.440 --> 11:04.510 So this is the same as here above 11:09.240 --> 11:14.880 and now we can also visualize the survival rate of the groups by applauding our pandas serious women 11:14.880 --> 11:21.840 and children first then we are using a bomb plot and we are setting the fixed size to 14 8 and the font 11:21.840 --> 11:26.280 size to 14 and let's run near this cell and let's see what we get here 11:29.260 --> 11:36.700 so we have here the battered Titanic survival rate by sex age groups and we can see here on the y axis 11:36.730 --> 11:43.780 we have these survival rate and the group with the highest survival rate is actually female adult then 11:43.780 --> 11:50.380 the second highest survival rate is female child and that by far lower survival rate has the adult male 11:50.380 --> 11:58.190 group here are right we are finished now with this video and I hope to see you in the next video by.