1 00:00:01,230 --> 00:00:06,550 Good we're continuing our series of lectures about manipulating Pandas data frames. 2 00:00:06,870 --> 00:00:12,030 So you learned how to drop a row or column out of a data frame in the previous lecture. 3 00:00:12,030 --> 00:00:17,170 Now let's see how we can add a column or row to a data frame. 4 00:00:17,670 --> 00:00:24,980 So we have df7 here with Address as index column and this series of column labels. 5 00:00:25,070 --> 00:00:26,720 Now just to clarify something. 6 00:00:27,030 --> 00:00:28,920 I know this is difficult to wrap up your mind. 7 00:00:28,920 --> 00:00:32,530 How will you be using these operations in real life? 8 00:00:32,580 --> 00:00:38,720 So I just want you to know with the syntax of doing these operations and throughout the course you learn 9 00:00:38,790 --> 00:00:43,180 how to actually put these into work with real life examples. 10 00:00:43,320 --> 00:00:44,840 So don't worry about that. 11 00:00:45,130 --> 00:00:47,120 And you know let's add a column there. 12 00:00:47,600 --> 00:00:55,790 To that you'd want to say df and in square brackets you pass the name of the new column. Let's say Continent. 13 00:00:56,500 --> 00:00:58,610 That would be equal to� 14 00:00:58,610 --> 00:01:04,330 You know now you would have to pass a list of values that you want to populate that column with. 15 00:01:04,350 --> 00:01:07,640 Let's say North America. 16 00:01:08,040 --> 00:01:11,940 Now if you execute that you'll get an error. Named df is not defined. 17 00:01:11,940 --> 00:01:14,740 Actually I didn't mean this error. 18 00:01:14,850 --> 00:01:17,210 This was another thing. You'll get this error. 19 00:01:17,460 --> 00:01:22,850 So it says that the length of values does not match the length of index. 20 00:01:23,700 --> 00:01:33,300 So the length of index is you know the length of the index is df7.index. 21 00:01:33,300 --> 00:01:34,520 The length of your index is five. 22 00:01:34,680 --> 00:01:39,270 But you're trying to pass there a list with a length of 1. 23 00:01:39,330 --> 00:01:41,210 So you've got only one element there. 24 00:01:41,820 --> 00:01:48,270 So the solution here is to pass a list with the exact number of items that you have in your table in 25 00:01:48,270 --> 00:01:49,860 your data frame. 26 00:01:49,890 --> 00:01:52,850 So we have five rows there five. 27 00:01:53,160 --> 00:01:58,630 And you know you could add four or more items here in North America. North America etc. 28 00:01:58,800 --> 00:02:10,600 Or you can do some fancy things in there. So you could say df7.shape and then zero times that. Execute. 29 00:02:10,870 --> 00:02:11,370 Df7. 30 00:02:11,380 --> 00:02:13,010 Print that out. 31 00:02:13,330 --> 00:02:14,840 Here we have a new column. 32 00:02:15,150 --> 00:02:17,190 So if you're confused with this shape. 33 00:02:17,310 --> 00:02:19,020 Zero times North America. 34 00:02:19,410 --> 00:02:27,660 Well what I did is you know df7.shape. What you get is 5,7 which means you have five rows and seven 35 00:02:27,660 --> 00:02:29,690 columns. 36 00:02:29,690 --> 00:02:34,970 Now what I want to get is the first item of that's tuple. 37 00:02:35,130 --> 00:02:42,190 So I get the five by doing shape zero. With this I always make sure that I'm getting the number of the rows 38 00:02:42,230 --> 00:02:44,170 that my data frame has. 39 00:02:44,880 --> 00:02:53,010 And then when you multiply five by a list with one element you get a list with five elements. 40 00:02:53,010 --> 00:02:53,700 41 00:02:53,700 --> 00:02:56,900 America. That�s the idea. 42 00:02:57,030 --> 00:03:05,660 Delete that and note this is actually an inplace operation so that will update your data frame and that 43 00:03:05,670 --> 00:03:07,370 was about adding a new column. 44 00:03:07,370 --> 00:03:09,670 And how about modifying a new column. 45 00:03:10,800 --> 00:03:13,800 So modifying the continent column. 46 00:03:14,010 --> 00:03:23,280 You know that could be something like country and you can also add some strings in there. Let's say plus 47 00:03:23,940 --> 00:03:28,140 comma plus maybe another string. 48 00:03:28,490 --> 00:03:29,330 North America. 49 00:03:32,100 --> 00:03:40,080 Execute that. Prints out the data frame and see what we got. 50 00:03:40,080 --> 00:03:46,380 So what we did here is we updated the Continent column by referring to an existing column which was 51 00:03:46,390 --> 00:03:47,670 Country. 52 00:03:47,900 --> 00:03:55,170 And then for each value of the country column we added so we concatenated the comma string which you can 53 00:03:55,170 --> 00:04:00,640 see in here just after USA and we also added another string. 54 00:04:01,230 --> 00:04:06,100 So North America. And this could also be another column if you liked. 55 00:04:06,330 --> 00:04:15,730 So if you pass here let's say Employees you'd get 8 instead of the first North America and then 15 56 00:04:15,740 --> 00:04:16,530 and so on. 57 00:04:16,590 --> 00:04:19,380 So that's how you update a column. 58 00:04:19,500 --> 00:04:21,370 How about adding a new row? 59 00:04:21,570 --> 00:04:25,650 Well this can be a bit tricky but still understandable. 60 00:04:26,160 --> 00:04:32,460 What you could do here because there is not an easy method to pass a row to a data frame� 61 00:04:32,760 --> 00:04:37,640 What you could do is you could say df7_t. 62 00:04:37,650 --> 00:04:45,010 So I'm creating a new variable that would be equal to df7.t. 63 00:04:45,030 --> 00:04:52,000 So T is actually is a method that what it does is it transposes your data frame. With transposition 64 00:04:52,020 --> 00:04:59,430 I mean you know df7, you check your new data frame and what you get is this. 65 00:04:59,580 --> 00:05:04,190 So your rows have become columns and your columns have become rows. 66 00:05:04,530 --> 00:05:13,320 So spend a few seconds looking at this you know. What we can do now is you know we can use the same syntax. 67 00:05:13,550 --> 00:05:20,820 DF7_t and then add a new column in there. So for the name of a column you will have 68 00:05:20,820 --> 00:05:23,260 to pass an address 69 00:05:23,750 --> 00:05:27,850 Let's say My address. 70 00:05:28,560 --> 00:05:37,370 That would equal again to the list and you want the list to reflect this order. So City First. 71 00:05:37,500 --> 00:05:42,000 Let's see My city. Country, My country. 72 00:05:45,040 --> 00:05:54,510 And 10 for employees and 7 for ID. My shop for the shop name. My state. 73 00:05:58,790 --> 00:06:05,990 And My continent. Execute thati. Df7_t. 74 00:06:06,090 --> 00:06:15,520 If you look at this now you'll see that you get a new column in your data frame and now what 75 00:06:15,550 --> 00:06:16,800 you do? 76 00:06:16,840 --> 00:06:29,690 So to complete the trick is let me say df7=df_t.T.. 77 00:06:29,970 --> 00:06:34,850 So df7 now will have the new row added at the end. 78 00:06:35,070 --> 00:06:36,940 So let me wrap this up. 79 00:06:36,990 --> 00:06:44,000 Again what I did is I transposed the original data frame df7 and than I added a column there. 80 00:06:44,140 --> 00:06:50,550 And if I transpose the data frame back again to its original position this column that I added will 81 00:06:50,550 --> 00:06:55,610 be converted to a row and that does the trick. 82 00:06:55,620 --> 00:06:59,780 Similarly you can modify or row if you like. 83 00:06:59,790 --> 00:07:07,110 So in that case you don't point to my address but you'd pointed to an existing column with an address 84 00:07:07,110 --> 00:07:08,730 with an existing address name. 85 00:07:08,910 --> 00:07:14,510 For instance this one if I pass it here and execute that. Execute that 86 00:07:14,540 --> 00:07:16,180 And that. 87 00:07:16,980 --> 00:07:23,730 And you'll see that the values of this row with this address name have been updated now. 88 00:07:23,740 --> 00:07:29,570 As you can see in here and that closes this lecture as well. 89 00:07:29,780 --> 00:07:31,050 And yeah, I'll see you later.