WEBVTT 00:01.000 --> 00:02.120 Hi everybody. 00:02.140 --> 00:08.200 In this video we analyze the pen that serious with non American value is so flexible String objects 00:08.380 --> 00:13.820 like names and therefore we work with the Summer Olympics data set. 00:13.990 --> 00:21.670 And first of all we import pandas and we import our Summer Olympic medal data set and we will have a 00:21.670 --> 00:30.400 look at the first five rows so you can see here each row is actually a medal that was awarded during 00:30.390 --> 00:31.840 a summer Olympic games. 00:32.110 --> 00:39.340 And here in the columns we have and the year of the Olympic Games and the host city the sport the discipline 00:39.880 --> 00:41.620 the athletes who won the medal. 00:41.620 --> 00:50.050 So here in this example if it hires won the gold medal and he's from Hungary and he's the man and the 00:50.050 --> 00:57.280 event was 100 meter freestyle and let's check also with the Inform method that we have in total thirty 00:57.280 --> 01:03.490 one thousand one hundred sixty five meters then that were awarded from eighteen hundred ninety six to 01:03.670 --> 01:13.250 2012 we have nine columns and we have more or less and no missing values only here in the country column 01:13.250 --> 01:20.780 for missing values and our year column contains in the US and all of the other columns contain actually 01:20.780 --> 01:28.370 String objects and now starting from our data frame here we can create a panel serious with the non 01:28.370 --> 01:37.140 numerical values so for example we slice for the column athlete and we assign the column athlete to 01:37.160 --> 01:38.720 the variable athlete. 01:39.230 --> 01:48.000 So let's do this and first of all we can check and the type of our column here and it's the panda series 01:48.760 --> 01:54.700 and let's have a look at the first five elements are rows of our panda serious. 01:54.700 --> 01:56.860 So here we have them the values. 01:56.860 --> 02:00.180 These are the athletes on the right side and on the left side. 02:00.280 --> 02:07.390 We still have an index of our pandas data frame and we can also have a look at the last five rows 02:10.470 --> 02:17.520 and we can also check the data type of them the values for the now pandas serious so to be here mastering 02:17.900 --> 02:19.780 we have year 0 4 object. 02:19.830 --> 02:26.160 So in simple words then we have an oh here it's very likely to String objects actually and we can also 02:26.160 --> 02:33.120 check the shape of our panda serious and we have a one dimensional array so a panda seriousness always 02:33.120 --> 02:42.130 one dimensional and we can also try to use and the described method and the described method gifts that's 02:42.220 --> 02:46.530 actually different results compared to a numerical series. 02:46.680 --> 02:54.830 So here for non numerical series we get first of all the total count of elements so we have thirty one 02:54.830 --> 03:01.610 thousand one hundred sixty five athletes and twenty two thousand seven hundred sixty two we are actually 03:01.610 --> 03:02.110 unique. 03:02.120 --> 03:05.860 So we have many athletes that won more than one medal. 03:05.870 --> 03:12.140 So for example Michael Phelps who is actually the top athlete who won the most medals and it's year 03:12.140 --> 03:13.130 frequency. 03:13.160 --> 03:16.550 You can see Michael Phelps 122 medals. 03:16.820 --> 03:20.060 And we can also check here at the size of our pandas serious. 03:20.060 --> 03:28.020 That's thirty one thousand one hundred sixty five and with the current method you can also check for 03:28.020 --> 03:29.830 the amount of normal values. 03:29.850 --> 03:30.370 And. 03:30.420 --> 03:33.780 This will see the same as with the size as attribute. 03:34.140 --> 03:40.650 So in our athlete column or panel series we do not have any missing or any end values. 03:40.650 --> 03:46.620 So if we had any end values and then the count method here what does that return a lower value than 03:46.620 --> 03:47.910 our size attribute. 03:48.420 --> 03:55.320 And the difference would then be actually the amount of missing values and we can also try to apply 03:55.380 --> 04:02.850 the min method so far in the map numerical serious we get to hear the minimum element and it's a non 04:02.850 --> 04:09.210 numerical values we get here as a better clear the very first athlete and that's the guy. 04:09.360 --> 04:15.010 I don't know how to pronounce it here but it starts with a double A B so alphabetically. 04:15.030 --> 04:22.960 This athlete or sexually the very first athlete and we can also check our pandas serious for unique 04:22.960 --> 04:32.620 values so we are all unique uh values are to say um athletes that won a medal in some Olympic Games 04:33.250 --> 04:40.060 and we can also count the number of unique athletes with the lungs function is twenty two thousand seven 04:40.060 --> 04:44.360 hundred sixty two and we can do the same with the method. 04:44.470 --> 04:50.770 Any unique sort of awesome returns us yet twenty two thousand seven hundred sixty two 04:53.460 --> 04:54.890 and it's unique. 04:54.900 --> 04:59.590 We can decide whether we want to skip and missing on a values or not. 04:59.820 --> 05:04.250 And we do not have any missing values here and our athletes series. 05:04.260 --> 05:06.780 So the result should be the same actually. 05:07.760 --> 05:09.310 So a default value is. 05:09.420 --> 05:10.080 True. 05:11.100 --> 05:18.210 And if we change to force so we should also look at the twenty two thousand seven hundred sixty two 05:18.210 --> 05:20.770 because we do not have missing values. 05:21.050 --> 05:22.240 All right. 05:22.380 --> 05:25.910 And last but not least we can also use them here. 05:25.960 --> 05:27.880 Father No no not in America. 05:27.880 --> 05:31.320 I'm serious and the very powerful method value Collins 05:34.440 --> 05:40.300 and what value Council actually does it returns us a panda serious with me actually. 05:40.310 --> 05:45.070 The frequency of all values some that appear in our pandas series so. 05:45.460 --> 05:51.290 Here we can see Michael Phelps is 22 times in our athlete pandas serious them. 05:51.300 --> 05:59.040 That means that Michael Phelps 122 Olympic medals in his career and value counts by default returns 05:59.070 --> 05:59.860 us here. 05:59.880 --> 06:02.200 The frequency and sort of order. 06:02.480 --> 06:08.100 So actually in descending order so the value with the highest frequency comes first here and that's 06:08.100 --> 06:16.690 Michael Phelps and the second most successful athlete a seer laterals are Latina and so on here then 06:16.800 --> 06:23.810 we can scroll down and in the end we end up with the athletes that won one gold medal here. 06:25.690 --> 06:30.790 And with the ascending perimeter we can also change the order of sorting. 06:31.330 --> 06:39.010 So here now we are starting with the athletes that won one medal and in the end we have here also Michael 06:39.010 --> 06:39.550 Phelps 06:45.580 --> 06:54.070 and we can also normalize our results to the 30 normalize perimeter so the normalized perimeter does 06:54.070 --> 06:58.080 not return absolute frequencies but relative frequencies. 06:58.210 --> 07:07.450 So we have some 31000 about medals and Michael Phelps won yeah only 0 point 0 0 7 0 6 percent of all 07:07.450 --> 07:08.650 medals. 07:08.660 --> 07:13.700 All right now we are finished with this session and the next session we will see and learn how to create 07:13.720 --> 07:14.990 panic serious. 07:15.160 --> 07:16.840 So hope to see you there by.