WEBVTT - This file was automatically generated by VIMEO 0 00:00:10.420 --> 00:00:12.680 Hi everyone, and welcome to our six module, 1 00:00:12.850 --> 00:00:15.960 which will be focused on data analysis in pandas. 2 00:00:17.750 --> 00:00:20.880 Okay. To introduce our module first, 3 00:00:21.540 --> 00:00:25.920 we will learn what is pandas data frame and what are, uh, 4 00:00:26.050 --> 00:00:26.883 panda series. 5 00:00:28.420 --> 00:00:33.410 We'll learn how to get initial information about them and how to work with the 6 00:00:33.410 --> 00:00:38.020 most common data audit, uh, methods. We'll learn, uh, 7 00:00:38.170 --> 00:00:42.660 what are pandas data types, which are supported and how to cast them. 8 00:00:44.490 --> 00:00:48.760 We'll learn about this index, the multi-index and how to work with it. 9 00:00:50.110 --> 00:00:54.540 We'll also learn the most common data frame operations, like, uh, 10 00:00:54.560 --> 00:00:58.740 how to create new column or rename columns, how to filter columns or roles, 11 00:00:59.560 --> 00:01:03.300 how to change values based on some condition and so on. 12 00:01:04.970 --> 00:01:09.360 We'll learn how to work with apply method. Also, 13 00:01:09.590 --> 00:01:14.290 very important is to know how to merge data frames. Also, 14 00:01:14.290 --> 00:01:17.530 how to summarize data with pivot table, 15 00:01:17.740 --> 00:01:21.620 cross up and using group. And lastly, 16 00:01:21.880 --> 00:01:25.500 we will learn how to style data frames in Jupyter Notebook. 17 00:01:27.400 --> 00:01:30.020 Now let's talk about the Pandas data frame and series. 18 00:01:31.230 --> 00:01:34.460 Let's go to our notebook in the module six folder. 19 00:01:36.450 --> 00:01:39.000 It'll be better to explain you all the concepts. 20 00:01:45.370 --> 00:01:49.670 So what is pandas data frame? It is the most important pandas class. 21 00:01:50.020 --> 00:01:53.510 Basically, it is a two dimensional data structure, 22 00:01:53.860 --> 00:01:58.710 like a matrix or a two dimensional array or a 23 00:01:58.710 --> 00:02:00.550 table with rows and columns. 24 00:02:02.090 --> 00:02:06.790 You can also say it is a higher level of wrapper around MPI arrays. 25 00:02:07.010 --> 00:02:11.750 And this higher level of functionality provides us easier way how to work with 26 00:02:11.750 --> 00:02:16.480 the data. It also consists of multiple panda series as well. 27 00:02:17.060 --> 00:02:21.520 Uh, basically if we filter only one column or one row, for example, 28 00:02:21.780 --> 00:02:23.360 we will get a panda series, 29 00:02:25.390 --> 00:02:28.640 both panda series and data frame share. 30 00:02:28.670 --> 00:02:31.000 Some of the methods which you can apply on them, 31 00:02:31.980 --> 00:02:35.360 and some are specific just to data frame or series. 32 00:02:37.140 --> 00:02:40.400 Now, how to create a pandas data frame in Python. 33 00:02:42.590 --> 00:02:47.010 Uh, the most common way how to do it, it is with, uh, from dictionary method, 34 00:02:48.680 --> 00:02:53.630 which creates a data frame from a Python dictionary where the keys 35 00:02:54.130 --> 00:02:58.470 are column names and the lists are the values for each of the columns. 36 00:02:59.870 --> 00:03:04.450 Uh, we can create a pandas data frame also from, uh, 37 00:03:04.980 --> 00:03:09.770 lists of dictionaries around this. So 38 00:03:16.930 --> 00:03:21.390 we can create a pandas data frame also from a list of dictionaries 39 00:03:22.040 --> 00:03:25.860 using from, uh, recourse method. Uh, 40 00:03:25.860 --> 00:03:30.820 where each item in the list is a dictionary where the key is the column name 41 00:03:32.200 --> 00:03:36.100 and the value is the value for this specific row and column. 42 00:03:40.300 --> 00:03:41.280 As I mentioned before, 43 00:03:41.820 --> 00:03:45.680 the data frame is a higher level wrapper around MPI array. 44 00:03:46.180 --> 00:03:50.120 We can get the MPI array with data frame values. 45 00:03:54.270 --> 00:03:54.490 Uh, 46 00:03:54.490 --> 00:03:59.050 to work with MPI arrays is more difficult than to work with panda data frame or 47 00:03:59.050 --> 00:04:02.410 series, but I recommend to you at least learn the basics. 48 00:04:04.210 --> 00:04:07.420 Each data frame has its, uh, index. Uh, 49 00:04:07.420 --> 00:04:11.060 these are the numbering of the rows. Here. 50 00:04:11.080 --> 00:04:13.780 We can see in bold and columns, 51 00:04:21.380 --> 00:04:25.680 uh, the index of the rows. Uh, it doesn't have to be just numbers, 52 00:04:25.900 --> 00:04:30.580 it might be strengths as well. We also see that when we refer, 53 00:04:30.760 --> 00:04:35.590 uh, to columns or the index, uh, we'll get a index class. 54 00:04:35.590 --> 00:04:39.440 In return, we can turn, uh, 55 00:04:39.440 --> 00:04:42.040 the index tool list with the tool list method. 56 00:04:48.670 --> 00:04:52.080 Okay, now what are Pandas series? Uh, 57 00:04:52.170 --> 00:04:56.480 panda series is uh, one dimensional array or one column, 58 00:04:56.780 --> 00:05:00.150 or one vector we can say as well. Uh, 59 00:05:00.150 --> 00:05:04.150 the easiest way how to create them is using, uh, cut. 60 00:05:05.660 --> 00:05:10.580 The easiest way how to create them is using a list. We can create them this way. 61 00:05:15.220 --> 00:05:18.790 Panda series can have a name, uh, 62 00:05:18.790 --> 00:05:21.150 same as column in a pandas data frame. 63 00:05:24.220 --> 00:05:26.400 It is also an index. We can see it here. 64 00:05:31.970 --> 00:05:36.390 And we see the difference that the name of the series is just, uh, one string, 65 00:05:37.650 --> 00:05:40.870 but the index is uh, also a range index class, 66 00:05:41.410 --> 00:05:43.190 and we can turn it into list as well. 67 00:05:49.840 --> 00:05:50.700 In our next video, 68 00:05:51.120 --> 00:05:55.020 we will learn how to load our data into Panta data frames.