Would you like to inspect the original subtitles? These are the user uploaded subtitles that are being translated: 1 00:00:01,050 --> 00:00:06,600 Hello everyone and welcome to the overview of data frame operations data frames are the workhorse of 2 00:00:06,600 --> 00:00:07,130 our. 3 00:00:07,350 --> 00:00:12,330 So in this lecture we're going to basically be going over all the common operations use of data frames 4 00:00:12,330 --> 00:00:15,090 and our this is going to be a very useful lecture. 5 00:00:15,100 --> 00:00:19,350 We're going to be going over material we've already covered since it's so vital to know data frames 6 00:00:19,470 --> 00:00:22,910 very well in order to save yourself time in the future. 7 00:00:22,920 --> 00:00:25,720 Let's go ahead and open up our studio and get started. 8 00:00:26,280 --> 00:00:33,840 OK so here our studio and you can see here in this the script I have listed a basically a list of topics 9 00:00:33,840 --> 00:00:37,380 we're going to be going over we're going to start off with creating data frames importing exporting 10 00:00:37,380 --> 00:00:42,990 data all the way down until we can get to how to deal of missing data and selecting multiple rows multiple 11 00:00:42,990 --> 00:00:43,950 columns. 12 00:00:43,950 --> 00:00:46,920 So we're going to take this one line slash topic at a time. 13 00:00:46,960 --> 00:00:51,650 You can always reference the notes from this lecture in case you interview anything we go over. 14 00:00:51,650 --> 00:00:56,460 We're going to go ahead and make the console larger and most of what we're going to be doing isn't in 15 00:00:56,460 --> 00:00:57,960 the con.. 16 00:00:58,050 --> 00:01:01,640 Let's go ahead and start with creating data frames. 17 00:01:01,690 --> 00:01:04,250 I in and clear the console of control here. 18 00:01:04,810 --> 00:01:05,640 OK. 19 00:01:05,940 --> 00:01:10,950 So if we want to just create a basic data frame just an empty data frame we can do something like this 20 00:01:10,950 --> 00:01:11,010 . 21 00:01:11,020 --> 00:01:16,800 You can say empty and data that frame is going to be the main function you can see our studios already 22 00:01:16,800 --> 00:01:18,300 completing it for us. 23 00:01:18,960 --> 00:01:22,530 And that's the way you can make just a basic empty data frame. 24 00:01:22,530 --> 00:01:27,250 Now let's go ahead and make a data frame from Vector we can say something like this. 25 00:01:27,340 --> 00:01:35,780 C1 which is going to be our first column is going to be just a vector of integers 1 through 10 reason 26 00:01:35,790 --> 00:01:43,740 that colon rotations quickly create that vector of integers and in other key trick to R is if you just 27 00:01:43,740 --> 00:01:49,980 type in letters letters is basically a vector of the alphabet that's built into our. 28 00:01:50,060 --> 00:01:53,390 You can quickly reference for letters A through Z. 29 00:01:53,400 --> 00:02:04,750 So what I'm going to go ahead and do is say C-2 and assign letters 1 through 10. 30 00:02:05,670 --> 00:02:08,830 Enough I check out C-One Anssi too. 31 00:02:09,120 --> 00:02:11,580 I have two columns that are available for me. 32 00:02:11,610 --> 00:02:15,650 Let's go ahead and combine those columns and make them part of a data frame. 33 00:02:16,040 --> 00:02:18,400 Then I go out and make an object called DSF. 34 00:02:18,540 --> 00:02:21,250 Use my keyword data frame data frame. 35 00:02:21,270 --> 00:02:24,950 That's going to be the function that allows me to build these data frames. 36 00:02:25,370 --> 00:02:37,750 I'll start off with call that name that one is equal to see one call not name that two is equal to C2 37 00:02:37,810 --> 00:02:38,770 . 38 00:02:39,840 --> 00:02:45,150 And then we can go ahead and check out this data frame if we can see here we have a rows and or two 39 00:02:45,150 --> 00:02:45,780 columns. 40 00:02:45,780 --> 00:02:49,590 So let's break down the state of frame function just pretty quickly. 41 00:02:49,620 --> 00:02:54,980 It's data frame and then it's the parameters that you pass in or the column names. 42 00:02:55,110 --> 00:02:57,630 And you can actually just create these call names you can make them up. 43 00:02:57,630 --> 00:02:59,360 They don't have to be a reference to something. 44 00:02:59,580 --> 00:03:04,530 And then you say equal and then what we actually want to pasan as the data for that column. 45 00:03:04,530 --> 00:03:07,490 So again I'm saying call that name too. 46 00:03:07,500 --> 00:03:11,460 So I want that to be the second column equals and then I pass in the data. 47 00:03:11,460 --> 00:03:16,450 If you don't pass in these variable names you're kind of making up for the column names. 48 00:03:16,500 --> 00:03:22,800 What are you going to do it's going to automatically name your columns for the variable names of whatever 49 00:03:22,800 --> 00:03:25,420 the data you were putting into it. 50 00:03:25,530 --> 00:03:28,900 All right now let's briefly go over importing and exporting data. 51 00:03:28,920 --> 00:03:32,570 There's a lot more lectures on imprinting exporting data to specific formats. 52 00:03:32,680 --> 00:03:37,740 Here we're just going to go have a brief overview of how to import and export from CSP files. 53 00:03:38,030 --> 00:03:39,680 Let's go at in clear a con.. 54 00:03:40,260 --> 00:03:44,170 To read a C as we filed the command the just read that C S V. 55 00:03:44,430 --> 00:03:48,810 That's going to be the most common command throughout the course as far as reading our information. 56 00:03:48,870 --> 00:03:55,080 They can say I've read that C S V and then you just pasan whatever your file name is you can say some 57 00:03:55,080 --> 00:04:03,660 file does CXXVI and this will go ahead and read that into a different if you want to writes to a C S 58 00:04:03,660 --> 00:04:04,180 V. 59 00:04:04,320 --> 00:04:08,550 You can just use write thought C S V passing your data. 60 00:04:08,550 --> 00:04:15,180 So for instance I can pass in the data frame we just made if and then the second argument is file and 61 00:04:15,180 --> 00:04:17,180 you pass in the file you want. 62 00:04:17,760 --> 00:04:22,910 So we can say something like Saved f not sci fi. 63 00:04:22,980 --> 00:04:25,590 So let's go ahead and read that CXXVI. 64 00:04:25,590 --> 00:04:30,950 So say DFI to read the CXXVIII. 65 00:04:30,960 --> 00:04:34,110 We'll go ahead and pass in the file name. 66 00:04:34,410 --> 00:04:40,470 So if we look at this it was saved underscore that CXXVI and I can actually press tab and it should 67 00:04:40,520 --> 00:04:49,430 auto complete to the filename you know we have DFI too which is the same thing as the if. 68 00:04:49,600 --> 00:04:57,790 But the thing to notice here is when you're saving it as a C S V with just these two parameters your 69 00:04:57,790 --> 00:05:01,650 index will also be saved as a column into that CAC file. 70 00:05:01,720 --> 00:05:05,200 So you'll notice that when we read that cxxviii file it's a Ze'ev too. 71 00:05:05,320 --> 00:05:11,570 We've got an extra column called X and the X column is just the index of that data frame that you saved 72 00:05:11,570 --> 00:05:12,970 at CAC. 73 00:05:13,420 --> 00:05:17,900 After reading this CACP you can either drop it keep it doesn't really matter what you do with it but 74 00:05:17,910 --> 00:05:21,820 just keep in mind that it's also going to save that index information. 75 00:05:21,880 --> 00:05:26,320 And the reason for that is because a lot of times in your data frames your index isn't just going to 76 00:05:26,320 --> 00:05:31,610 be numbers are actually going to be string names such as names of states or people cetera. 77 00:05:31,690 --> 00:05:36,400 So it's good to say that information when you're writing to see every file when you're reading it you 78 00:05:36,400 --> 00:05:41,470 can go ahead and either delete drop or do whatever you want that information. 79 00:05:41,470 --> 00:05:42,040 All right. 80 00:05:42,040 --> 00:05:47,650 Now let's go ahead and move on to our third topic getting information about the data frame going to 81 00:05:47,650 --> 00:05:49,040 go ahead and clear the con.. 82 00:05:49,540 --> 00:05:51,870 And we have our data frame Dia. 83 00:05:52,460 --> 00:05:54,920 Let's who wanted to know the row and column counts. 84 00:05:54,940 --> 00:06:00,880 There's a couple of ways to do this but the most direct way is the NRO function that's all lowercase 85 00:06:00,940 --> 00:06:01,050 . 86 00:06:01,090 --> 00:06:03,430 And then you just go ahead and pass in your data frame. 87 00:06:03,730 --> 00:06:10,060 And I'll tell you back how many rows you have and you can also say and cnl for end columns passing data 88 00:06:10,060 --> 00:06:12,110 frame it'll tell you how many columns you have. 89 00:06:12,110 --> 00:06:16,980 Back again this doesn't include the actual index just the columns that have names. 90 00:06:17,350 --> 00:06:25,300 If you actually want to get those names you can use the call names function so that C O L names and 91 00:06:25,300 --> 00:06:30,390 then just call on any data frame and all or turnback a vector of the column names. 92 00:06:30,490 --> 00:06:32,110 You can do the same with forenames. 93 00:06:32,120 --> 00:06:36,200 However we should probably be careful of this because we have a very large data frame. 94 00:06:36,470 --> 00:06:38,460 This output will be really large. 95 00:06:38,470 --> 00:06:39,660 So just keep that in mind. 96 00:06:39,730 --> 00:06:44,590 And in this case where that data frame just has an index it's not super useful to have the road names 97 00:06:44,830 --> 00:06:45,560 . 98 00:06:45,670 --> 00:06:51,130 Couple of other ways to get information but data frame is the SDR function which returns the structure 99 00:06:51,190 --> 00:06:52,040 of the data frame. 100 00:06:52,090 --> 00:06:58,210 So here are the math here I'll tell you how many observations you have of how many variables which are 101 00:06:58,210 --> 00:07:01,440 the same as just those columns will tell you the column names. 102 00:07:01,690 --> 00:07:07,940 So you kind of the data type is in those columns so integers or a factor of 10 levels etc.. 103 00:07:08,110 --> 00:07:10,600 That's one way of getting information off the data frame. 104 00:07:10,870 --> 00:07:15,820 The other way is with summary and this will give you a statistical summary of what's going on in your 105 00:07:15,820 --> 00:07:17,260 data frame. 106 00:07:17,260 --> 00:07:22,040 Some of these don't make too much sense if you're looking at it from a statistical standpoint such as 107 00:07:22,040 --> 00:07:24,450 a column name to the column name one. 108 00:07:24,450 --> 00:07:27,520 If we check a look it's just a bunch of integers. 109 00:07:27,730 --> 00:07:29,910 So here we get statistical information on the median. 110 00:07:29,910 --> 00:07:36,190 The mean the max value cetera recall him to since it was a bunch of string values what you're going 111 00:07:36,190 --> 00:07:42,100 to get in return instead of the sort of statistical information it's just a count of the terms how many 112 00:07:42,100 --> 00:07:42,940 times they show up. 113 00:07:42,940 --> 00:07:44,140 ABC the. 114 00:07:44,230 --> 00:07:44,880 Cetera. 115 00:07:45,010 --> 00:07:49,920 And if there's too many values here they'll eventually say other and give you another count. 116 00:07:49,960 --> 00:07:52,650 Otherwise this would be a really long output. 117 00:07:53,140 --> 00:07:53,970 OK. 118 00:07:54,280 --> 00:07:56,650 That's it for getting information about a data frame. 119 00:07:56,650 --> 00:07:57,710 Our next topic. 120 00:07:57,790 --> 00:08:00,790 Topic number four is going to be referencing cells. 121 00:08:00,790 --> 00:08:06,070 So now we're going to learn how to reference a particular cell off a data frame and go out and clear 122 00:08:06,070 --> 00:08:06,940 this. 123 00:08:06,940 --> 00:08:08,730 Check out our data from DIA. 124 00:08:09,340 --> 00:08:15,970 And there it is the most basic way to reference a single cell in a data frame is by using two sets of 125 00:08:15,970 --> 00:08:17,980 brackets with index numbers. 126 00:08:17,980 --> 00:08:22,010 For example you could say if your first set of brackets. 127 00:08:22,120 --> 00:08:27,710 And then inside a second set of brackets you'll pasan a row comma column value. 128 00:08:27,730 --> 00:08:35,200 So imagine you wanted to get the value at row number five on column two or I could say something like 129 00:08:35,200 --> 00:08:38,260 this 5 which is the row column too. 130 00:08:38,470 --> 00:08:44,380 So if I look at that I get back e as well of some more information on the levels where the levels is 131 00:08:44,380 --> 00:08:48,620 basically just unique values in this case of column name too. 132 00:08:48,670 --> 00:08:50,010 So it's actually going on here. 133 00:08:50,050 --> 00:08:55,000 We checked that row five value in column two which is. 134 00:08:55,390 --> 00:08:59,920 Now this isn't going to be super useful because a lot of times you want to reference the actual column 135 00:08:59,980 --> 00:09:02,560 name not the column number. 136 00:09:02,860 --> 00:09:07,180 So the way you do that is really similar pasan a set of brackets. 137 00:09:07,180 --> 00:09:09,940 Again the row you want in this case it's five. 138 00:09:09,940 --> 00:09:16,090 You can also pass on that index name and in the string you can pass in the column name instead of the 139 00:09:16,090 --> 00:09:17,580 column number. 140 00:09:17,720 --> 00:09:22,330 So call name 2 here and we get back the same value. 141 00:09:22,330 --> 00:09:27,970 You're probably going to be using this sort of reference usually instead of memorizing what positions 142 00:09:27,970 --> 00:09:34,600 the columns are in and we can use this sort of syntax to actually reassign single cell values. 143 00:09:34,770 --> 00:09:40,380 So for instance let's say I wanted to change this to into a number like nine thousand nine hundred ninety 144 00:09:40,380 --> 00:09:41,170 nine. 145 00:09:41,460 --> 00:09:45,900 I would say DMF has a second set of brackets. 146 00:09:46,020 --> 00:09:57,240 That too is in rote too and it's in call that name one and I'm going to head and use a assignments to 147 00:09:57,250 --> 00:09:58,720 pass in a number there. 148 00:09:58,750 --> 00:10:03,100 So let's go in and say nine thousand nine hundred ninety nine that does the assignment. 149 00:10:03,120 --> 00:10:04,830 So if I take a look at. 150 00:10:05,280 --> 00:10:08,620 Notice now that value is nine thousand nine hundred ninety nine. 151 00:10:08,730 --> 00:10:14,040 So I can call that single cell value and then pass an assignment to it. 152 00:10:14,070 --> 00:10:19,500 So that's an important feature to know that you can use when you're referencing single cells. 153 00:10:19,500 --> 00:10:19,870 All right. 154 00:10:19,890 --> 00:10:26,040 Let's go ahead and move on to referencing rows so Rose is going to be similar except without that double 155 00:10:26,040 --> 00:10:33,030 bracket notation going to go and clear this in order to return a row which you can do is something like 156 00:10:33,030 --> 00:10:33,270 this. 157 00:10:33,270 --> 00:10:41,790 You can say DSF racquet pass in the number or index value of the row you want comma nothing. 158 00:10:41,790 --> 00:10:45,970 Now we should be pretty familiar with this sort of format from our study of Meecher season. 159 00:10:46,830 --> 00:10:47,890 And there it is. 160 00:10:47,940 --> 00:10:48,650 That's the row. 161 00:10:48,660 --> 00:10:51,150 And it returns a data frame not a vector. 162 00:10:51,150 --> 00:10:52,460 So it's something to keep in mind. 163 00:10:52,590 --> 00:10:55,820 You'll still have information of those two column names. 164 00:10:56,760 --> 00:11:01,770 If for some reason you actually didn't want that data frame information you just wanted a straight vector 165 00:11:02,160 --> 00:11:07,920 of those values which you could do is something like an as numeric call to try to convert that row to 166 00:11:07,920 --> 00:11:09,460 a vector. 167 00:11:09,720 --> 00:11:14,960 But again the easiest way to get back a row is just called data frame a single set of brackets. 168 00:11:15,120 --> 00:11:20,250 The row you want comma nothing and then close that bracket. 169 00:11:20,250 --> 00:11:24,960 Let's talk about referencing columns which is going to be kind of similar except our placement is going 170 00:11:24,960 --> 00:11:26,400 to be a little different. 171 00:11:27,060 --> 00:11:32,370 Let's go ahead and play around with the empty cars state a frame that comes built into our. 172 00:11:32,490 --> 00:11:37,290 So the empty car as we take a look at it again it's a bunch of cars and then a bunch of columns with 173 00:11:37,300 --> 00:11:38,210 their difference. 174 00:11:38,250 --> 00:11:43,400 Are there stats like MPG number cylinders horsepower etc.. 175 00:11:43,740 --> 00:11:46,810 Let me go ahead and change to suggest looking at the whips. 176 00:11:47,040 --> 00:11:51,000 The head of cars. 177 00:11:51,450 --> 00:11:51,750 All right. 178 00:11:51,880 --> 00:11:55,240 So that's a little more manageable and a control. 179 00:11:55,260 --> 00:11:57,030 I'll do that one more time. 180 00:11:57,090 --> 00:11:58,440 So we have a cleaner screen. 181 00:11:58,800 --> 00:12:00,760 OK we're looking at the head of Indy cars. 182 00:12:00,880 --> 00:12:03,360 There's lots of different ways to actually get a column. 183 00:12:03,370 --> 00:12:06,550 Let's go ahead and talk about four main ways. 184 00:12:06,900 --> 00:12:15,690 One way is just to get a vector back so I can say cars or excuse me empty cars dollar sign and then 185 00:12:15,690 --> 00:12:22,380 the name of my column says one of the most common ways to grab columns as a vector of values. 186 00:12:22,500 --> 00:12:28,260 And you'll notice this is common enough since our studio will actually have a nice little dropdown menu 187 00:12:28,260 --> 00:12:29,970 of all the columns available for you. 188 00:12:29,980 --> 00:12:34,650 So we'll go on to say NPG and then we get back a vector of all the values. 189 00:12:34,650 --> 00:12:40,900 One other way to do this is going to be similar to the way we reference the row. 190 00:12:41,140 --> 00:12:47,730 It's just essentially going to be backwards so you can say nothing comma and then pass in mpg and you'll 191 00:12:47,730 --> 00:12:49,980 get back a vector of those values. 192 00:12:49,980 --> 00:12:55,190 You could have also passed in the number of that column. 193 00:12:55,260 --> 00:12:57,180 So MPG is the first column. 194 00:12:57,180 --> 00:13:02,490 So if I pass then a comma one there I would have also gone back mpg. 195 00:13:02,490 --> 00:13:04,460 And finally a little less common. 196 00:13:04,470 --> 00:13:10,720 But another way we could add this is passing in two sets of brackets like this with a single column 197 00:13:10,710 --> 00:13:12,930 name such as mpg. 198 00:13:13,590 --> 00:13:19,530 OK so those are four different ways you can get back a vector of those column values so empty cars mpg 199 00:13:20,490 --> 00:13:25,450 just with the dollar sign a comma with nothing in front of it. 200 00:13:25,680 --> 00:13:32,190 First comma and then either the name of the column you want or it's position in relation to the other 201 00:13:32,190 --> 00:13:33,780 columns. 202 00:13:33,780 --> 00:13:37,430 Or you can just use a double set of brackets of the column name. 203 00:13:37,470 --> 00:13:43,310 Now something to quickly note here is this were methods of just returning the vectors if we wanted to 204 00:13:43,320 --> 00:13:45,030 actually return a data frame. 205 00:13:45,030 --> 00:13:49,310 So keep this car information here is the MPG with it. 206 00:13:49,360 --> 00:13:58,540 We would just use two other methods so that we could say empty cars single set of brackets mpg and that's 207 00:13:58,530 --> 00:14:02,940 going to return a data frame instead of just a vector of the values. 208 00:14:02,940 --> 00:14:07,680 So note the difference here between these two methods using two sets of brackets is going to return 209 00:14:07,680 --> 00:14:09,340 that vector of values. 210 00:14:09,510 --> 00:14:15,320 If we just use a single set of brackets they'll return a data frame with just that column back. 211 00:14:15,690 --> 00:14:21,100 Now we could have also instead of having to say mpg if we knew the order of the columns we could always 212 00:14:21,100 --> 00:14:25,920 just replace this with an integer value for that column which in this case is one that said use the 213 00:14:25,920 --> 00:14:27,100 first column. 214 00:14:27,610 --> 00:14:33,000 So keep in mind you have this method available to you if you want to add data frame with that column 215 00:14:33,000 --> 00:14:33,940 back. 216 00:14:34,410 --> 00:14:38,270 Let's go ahead and clear the con.. 217 00:14:38,600 --> 00:14:42,660 And last thing I want to talk about before we move on from referencing columns is how to get multiple 218 00:14:42,660 --> 00:14:45,120 columns back as a data frame. 219 00:14:45,120 --> 00:14:53,720 So if we wanted to for instance get empty cars this would just return the MPG column who want multiple 220 00:14:53,730 --> 00:14:59,460 columns we could pass any vector of column names such as MPG and cylinders. 221 00:14:59,460 --> 00:15:04,560 Let's say that I'm going to go ahead and type and head at the beginning of all this. 222 00:15:04,560 --> 00:15:07,380 So you only see the first six rows. 223 00:15:08,010 --> 00:15:13,380 Whoops forgot to close off the vector. 224 00:15:13,410 --> 00:15:14,370 There it is. 225 00:15:14,880 --> 00:15:15,470 OK. 226 00:15:15,690 --> 00:15:21,000 So again empty cars instead of just passing MPG to get a single column back as the data frame. 227 00:15:21,050 --> 00:15:25,470 Can pass on a vector of column values you want and you can also pass them back in in whatever order 228 00:15:25,470 --> 00:15:25,850 you want. 229 00:15:25,870 --> 00:15:29,880 They don't have to be in the same order they are in that original data frame. 230 00:15:29,880 --> 00:15:30,150 All right. 231 00:15:30,150 --> 00:15:35,520 So you've gone over creating data frames importing exporting data from a CNC file getting information 232 00:15:35,520 --> 00:15:39,710 about the data frame referencing cells rows and columns. 233 00:15:40,020 --> 00:15:41,310 We'll go ahead and stop here. 234 00:15:41,350 --> 00:15:45,660 And then the next lecture will have a part two to this lecture we'll all go over adding rows adding 235 00:15:45,660 --> 00:15:51,090 columns setting column names or multiple rows or multiple columns and then dealing with missing data 236 00:15:51,100 --> 00:15:51,820 . 237 00:15:51,810 --> 00:15:56,190 All right thanks everyone and I'll see you at the next lecture which is just going to be part two of 238 00:15:56,250 --> 00:15:57,960 overview data frame operations 25024