WEBVTT 00:01.370 --> 00:08.330 In this situation, we would learn about Fondas and we will learn how we can work with him, so we would 00:08.330 --> 00:15.970 implement what we have learned to critical detail now and what we have learned in pieces in Fondas. 00:16.850 --> 00:18.300 So let us begin with that. 00:18.800 --> 00:24.710 So first of all, we'll import numbers and find us after importing Memphian Fondas. 00:24.710 --> 00:31.640 I'm just creating a new Dufrene with each physical safety I.D. rating, violence and children as columns 00:32.210 --> 00:33.580 and these other details. 00:33.590 --> 00:43.210 Now you can see that the age has values which are numeric in nature and it has strong values as defined. 00:43.220 --> 00:49.940 And thirty four, which is again, no, but the data is strong and apart from that that is of the string 00:50.300 --> 00:53.030 which is missing, which is present in each. 00:53.690 --> 01:01.130 Then we have FICO range, which is entirely in a string form but is created out of numbers and no hyphen. 01:02.260 --> 01:03.940 Apart from that, we have cities. 01:05.730 --> 01:06.000 Mea. 01:07.370 --> 01:12.620 Rioting and violent violence also contains not a number. 01:13.690 --> 01:18.800 And children, Hasle range from zero to five. 01:19.240 --> 01:26.560 And we have created 50 such rules, so I just run this and create a data frame using the dictionary 01:26.560 --> 01:28.540 for my annual printing. 01:28.540 --> 01:35.080 The head of the data frame, you can see we have these particular values and you can see in each column 01:35.080 --> 01:40.570 we are not able to find any missing values as of now. 01:40.780 --> 01:45.070 So let me just extend this a little bit and see if these 15. 01:45.880 --> 01:56.140 So when I see 15 rows, yes, there are missing values and rule 11 and rule index 40 and there is an 01:56.200 --> 01:58.780 order, no value and balance. 01:59.700 --> 02:07.050 In Rule nine, so these are a few values which we will need to keep in consideration and try to modify 02:07.050 --> 02:08.040 and try to in. 02:09.070 --> 02:15.730 Now, the next step, which we have is I can take the data types, so when I see my detailed or Dee 02:15.730 --> 02:23.050 Daibes, so I will be getting the data out of the columns which I have or the features or attributes 02:23.050 --> 02:23.750 that I have. 02:24.480 --> 02:26.420 Now, what are these other groups? 02:26.440 --> 02:30.180 It is Aidy now Ibbs and Dijo, which is perfectly fine. 02:30.850 --> 02:39.260 Age is object, which means that each is a string but each has to be a numeric because each is a number. 02:39.260 --> 02:39.640 Right. 02:39.790 --> 02:42.090 And here all these values are numbers. 02:42.250 --> 02:44.770 So we have to convert each into a number. 02:45.040 --> 02:49.660 That is the first point which we got from the second point is FICO. 02:49.840 --> 02:52.170 FICO is a numeric type. 02:52.180 --> 02:54.310 It is a data string. 02:54.490 --> 03:00.490 But if the string because there is a range, there is a hyphen in between and there are two numbers 03:00.730 --> 03:01.210 in here. 03:01.540 --> 03:11.940 Now, we cannot really apply any mathematical operation on a string and that to having two numbers together. 03:12.400 --> 03:18.160 So if you want to apply a mathematical operation, we need to have the hundred separate and one for 03:18.160 --> 03:25.660 three separate so that I can either multiply with some number or divided with some number or do any 03:25.660 --> 03:27.320 mathematical operation on top of it. 03:27.670 --> 03:30.610 So the next task which I have in hand is. 03:31.630 --> 03:41.320 To split these, so I want to have one hundred and one fifty in two different columns, or if I don't 03:41.320 --> 03:49.090 create two different columns out of what I can do is I can take the mean of the physical value and create 03:49.090 --> 03:51.160 a new column with only the mean value. 03:51.520 --> 03:57.790 Because you see the range for all of these are 50, 150, minus hundred is fifty three hundred minus 03:57.790 --> 04:01.180 two fifty fifty to fifty minus two hundred and fifty. 04:01.450 --> 04:05.020 So all the differences between these values is 50. 04:05.290 --> 04:06.600 So the range is fixed. 04:06.820 --> 04:14.320 So what we can do is we can have a mean value and create a column which is just for me I'm in gives. 04:14.320 --> 04:15.550 We don't want coming. 04:15.550 --> 04:19.030 We can create two columns as we come in and go, Max. 04:19.240 --> 04:24.040 So this is completely our choice, how we want to modify this particular column. 04:25.590 --> 04:32.880 The next column is City Now, city is a string of values, is having strong values, so it is a categorical 04:32.880 --> 04:33.210 column. 04:33.520 --> 04:39.540 Now, what we have known for that to the column is that we will be creating dummies out of this. 04:40.360 --> 04:45.930 That is, we will be creating one encoding, which is basically we will be creating a zero one zero 04:45.930 --> 04:46.580 one nothing. 04:46.800 --> 04:52.190 So we will have a column name Chinni and below zero or one values. 04:52.200 --> 04:55.230 Based on that, there is a value present or not. 04:58.230 --> 05:06.480 So we can work my way into another column on that column again, we'll have zero and one values depending 05:06.480 --> 05:09.900 on if the value is Mumbai for that particular. 05:10.410 --> 05:16.750 So each role will be converted into instead of a one city column. 05:16.770 --> 05:23.490 It will be converted into four columns with each of the values of Jenay Daily Woombye and. 05:24.790 --> 05:31.870 So we'll see how we do that next is having the reading column, this reading column will have again, 05:31.870 --> 05:34.720 excellent vibe, excellent, but I think good values. 05:34.870 --> 05:36.040 So these values. 05:38.140 --> 05:38.740 Have. 05:40.040 --> 05:41.570 US order in prison. 05:42.110 --> 05:46.670 So there is an order present in these values, so these are ordinary values. 05:46.850 --> 05:51.120 So ratings because it is the ordinary column or the new variable. 05:51.350 --> 05:55.290 So we will convert this froma string type to one number nine. 05:55.700 --> 06:02.120 We will give specific numbers to good, pathetic, excellent buy based on the sequence which is present 06:02.360 --> 06:07.000 so we can give BIID as one, then pathetic as zero. 06:07.160 --> 06:14.720 We can give good to let us see two and then excellent two three with three being the best rating and 06:14.720 --> 06:17.050 zero being the most reaping now. 06:17.300 --> 06:20.330 Next thing is balance, which is a floating coin. 06:20.340 --> 06:21.080 So this is fine. 06:21.080 --> 06:26.210 So balance has a floating point value, so it does not need to be converted into a numeric form. 06:26.360 --> 06:31.700 Similarly, children is again in digital form, so it does not have to be converted into a numeric form, 06:31.700 --> 06:33.120 which is already the numerical. 06:33.510 --> 06:41.750 Now, as we look at these values now, age has these missing values and balance also has these not a 06:41.750 --> 06:47.580 number of values, which means that we will have to imbue these values here. 06:48.680 --> 06:55.820 So what we do is for these data types, once we have checked this, we will start modifying these values 06:55.820 --> 06:56.460 one by one. 06:56.720 --> 07:03.700 So here, if you see for my data age, what I am doing here is I am imputing the values. 07:03.720 --> 07:06.620 I am just converting these values into a numeric form. 07:06.620 --> 07:07.840 Now, how do I do it? 07:08.060 --> 07:11.060 I do it with the user function to numeric. 07:12.020 --> 07:16.190 So there's two numeric function, basically convert this. 07:17.360 --> 07:28.640 A mixture of strings and numeric type in a B or numeric data and converting that, let us see if I just 07:28.640 --> 07:31.580 simply do it this way, I just remove this. 07:31.910 --> 07:35.330 It is not good cause because we don't know what that is as of now. 07:35.510 --> 07:42.410 So once I run this, what happens is you can see that it gives me an error that unable to string missing 07:42.410 --> 07:43.160 and position. 07:43.820 --> 07:45.620 So I position 11. 07:45.620 --> 07:51.740 It found a string which is missing and it is not able to understand what number should it please instead 07:51.740 --> 07:52.900 of the string missing. 07:53.180 --> 07:55.850 So it is not able to convert it properly. 07:56.060 --> 08:05.510 So to resolve this issue, we basically have this error cause what it does is whenever it finds error, 08:05.660 --> 08:08.290 it will convert that into not a number. 08:09.250 --> 08:17.140 So I'll just run this for you, so when I run this, it has converted these values into Northern No, 08:17.140 --> 08:19.390 you can see at 11 they have an odd number. 08:19.390 --> 08:21.160 At 14, they have the number. 08:21.430 --> 08:23.940 And if you will, check the data type again. 08:24.190 --> 08:25.780 So let me run this one again. 08:25.790 --> 08:31.000 So if you see the database back in the day that I would have changed to number. 08:33.350 --> 08:41.090 Let me do this again, so here you see the leadership has changed to a floating point, so we have resolved 08:41.090 --> 08:42.180 this particular issue. 08:42.680 --> 08:50.540 Now, next, what we can see here is we have we can perform different operations like the new column, 08:50.750 --> 08:54.450 or we can generate columns using the existing columns. 08:54.680 --> 08:58.810 So here what I'm doing is I'm creating a column named Constant Variable. 08:59.220 --> 09:02.030 I'm giving the value one hundred and say this. 09:02.330 --> 09:03.380 So I'll run this. 09:03.590 --> 09:06.800 And when I run this, you can see the data. 09:07.100 --> 09:11.020 So I have created a constant variable which has value. 09:11.040 --> 09:18.380 Hundert, how we create this is I just give the variable name, the column name in the races which I 09:18.380 --> 09:20.600 want to create and the value in front of it. 09:20.750 --> 09:23.390 So it has created this constant variable with value. 09:24.290 --> 09:28.280 Next, I give a new column name, which is violence. 09:28.280 --> 09:29.410 Log on. 09:29.600 --> 09:37.250 The formula for it is and we don't look, I'm just applying the mathematical function log to the violence 09:37.250 --> 09:40.070 value so it picks out the violence value. 09:41.830 --> 09:46.300 And big a logo and puts it in the violence column. 09:46.610 --> 09:54.250 Now, another thing is wherever the value of the violence would be, not a number, the value of violence 09:54.520 --> 09:55.420 also comes out. 09:55.450 --> 09:56.390 Do we know the number? 09:56.770 --> 10:04.870 So once we are doing such operations, so what we will have to do is if we are imputing the values to 10:04.870 --> 10:12.220 the violence column, we will also have to invoke new values to the violence column or what we can do 10:12.220 --> 10:18.920 is be reinforced if we import the values in the violence column and then generate the violence column 10:18.940 --> 10:22.480 so that it will automatically get the values from the violence column. 10:23.770 --> 10:30.880 Now, the next thing is maybe the age children ratio now here what we are simply doing is we are getting 10:30.880 --> 10:37.930 the age data and the children data and we're just dividing the data by the children data so that we 10:37.930 --> 10:41.050 could get the ratio of now. 10:42.480 --> 10:47.760 This is, again, the same thing, so because it has a number here. 10:48.970 --> 10:54.840 So I know the number has been generated, so these are a few things which we need to handle accordingly. 10:57.850 --> 11:06.760 Now we can do as many complex calculations, we want to create new columns and wherever we have missing 11:06.760 --> 11:09.370 values, we will have to imbue those values. 11:09.520 --> 11:12.790 Now, how we will impose those values will look at that little. 11:14.040 --> 11:22.150 Here we have another column maybe that each now we will check if there is any value in this each column. 11:22.560 --> 11:28.780 Now, as we already know, there are a lot of northern values in this particular each column. 11:29.070 --> 11:36.570 So when we do some of these so I'll just run this so you can see whatever the value was, not that we 11:36.570 --> 11:37.530 are getting a true. 11:39.700 --> 11:46.160 And when they run up some more food, so it will give me some of those values. 11:46.330 --> 11:49.450 So in total there are five nonsuch values. 11:50.020 --> 11:57.730 So what we can do is we need to impute values that we need to fill these places so that whenever we 11:57.730 --> 12:03.400 are performing any mathematical operation or whenever we are creating a model on top of it, it does 12:03.400 --> 12:09.460 not have any values because in case they would be null value, then it would not be able to process 12:09.460 --> 12:11.670 that to train our models. 12:12.310 --> 12:20.830 So whenever we are training on to make sure that no column has any knowledge, values or any missing 12:20.860 --> 12:21.430 values. 12:22.900 --> 12:24.400 So what we do here is. 12:26.130 --> 12:35.730 For my data, not location, so I'm filtering out the locations in my my data data frame and what am 12:35.730 --> 12:35.840 I? 12:35.930 --> 12:39.570 So I'm filtering my B that H. 12:40.590 --> 12:49.680 So I'm filtering in the my data age column and what am I taking, I am checking wherever it is. 12:50.920 --> 12:56.440 So this is basically filtering out the rules, maybe the. 13:01.940 --> 13:07.780 So this is checking their my data, each column has a value. 13:08.150 --> 13:13.460 It is just filtering out those five rows for me and. 13:14.860 --> 13:19.850 It is taking out the each column out of it, so we are simply getting these values. 13:20.050 --> 13:22.690 So let me run only this much piece of quote. 13:28.860 --> 13:35.310 So here you can see we have figured out only the values, which are not the number in the each column. 13:35.490 --> 13:41.390 So now what we are doing is we are simply pushing in my data each, not me. 13:41.400 --> 13:48.150 So we are taking the mean of the mighty that each column and we are pushing it inside these five rows 13:48.150 --> 13:48.660 of data. 13:49.570 --> 13:56.530 Similarly, what I'm doing here is I'm taking the mighty the eight children ratio and I'm updating it 13:56.740 --> 14:01.040 with the updated my data age and my little children. 14:01.180 --> 14:07.420 So now there will not be any null values which would have been imputed earlier because of the knowledge 14:07.420 --> 14:07.970 values. 14:08.920 --> 14:12.250 So when I ran this and then done this again. 14:12.400 --> 14:16.250 So now you can see there is no null value present in this particular column. 14:16.280 --> 14:22.890 And when I run my data again, so you can see that the values which were not that is the 11th rule. 14:23.740 --> 14:24.760 I'm the 14. 14:24.780 --> 14:33.080 True, they have been computed with the mean of the each column I'm the same has been updated in the 14:33.170 --> 14:33.610 children. 14:33.610 --> 14:33.970 Rachel. 14:36.170 --> 14:40.580 Now, the next thing what we can do is so here we have the reading school. 14:40.650 --> 14:43.460 So now reading was or the news column. 14:43.760 --> 14:45.040 So what did we decide? 14:45.050 --> 14:48.730 We will give some reading to the ordinary values. 14:48.740 --> 14:55.640 So instead of creating the values, what we do for all the variables is that we are pushing different 14:55.640 --> 14:57.640 values to the goal itself. 14:58.160 --> 15:05.600 So what I'm doing here is in the pathetic column, I'm pushing it minus one in the bag column I'm pushing 15:05.600 --> 15:15.500 in Zettl in the good or excellent column we are pushing in the one volume so that the value for reading 15:16.010 --> 15:17.140 is pathetic. 15:17.330 --> 15:18.800 It will be replaced by one. 15:19.130 --> 15:22.430 Whenever it is five, it will be replaced by zero. 15:22.640 --> 15:25.850 Wherever it is good or excellent, it will be replaced by one. 15:26.720 --> 15:28.310 So now what I do here is. 15:29.180 --> 15:31.150 I check my little reading school. 15:32.120 --> 15:39.380 This is a new column which I'm creating my writing school and I'm checking and beyond with this is a 15:39.380 --> 15:44.650 new format for you and we don't where it is checking a particular location. 15:44.930 --> 15:48.740 So where they're checking it is checking into my data. 15:48.740 --> 15:52.850 Reading it is checking values in the middle eating. 15:53.070 --> 15:54.290 Now, what is it checking? 15:54.290 --> 15:59.330 It is checking that the value inside my eating is in. 15:59.540 --> 16:07.280 If the value in the military thing is a part of this particular list, if any of these values is a part 16:07.280 --> 16:10.670 of good or excellent, then or one otherwise make a. 16:12.370 --> 16:19.150 So wherever the valley would be good or excellent, it would be changed to one and all of the values 16:19.150 --> 16:19.920 would be converted. 16:20.590 --> 16:26.500 Now see, whatever the value is good or excellent it has been converted to, one would just suffice. 16:26.840 --> 16:31.900 Now, the other situation was the buyer has to be converted to zero, which is also fine. 16:31.900 --> 16:36.040 That also has been done by putting Zettl in the other condition. 16:36.220 --> 16:41.800 Now, the only thing that we need to make sure is that whatever the value is for Turkey, that has to 16:41.800 --> 16:43.270 be replaced by minus one. 16:43.450 --> 16:50.080 So what we do is in my location, my be eating whatever the value is, but typic. 16:51.610 --> 16:55.980 We are opening the values for ratings, so we are getting the values for ratings. 16:56.430 --> 17:00.080 So what we do is, again, we will run only this much piece of gold. 17:03.050 --> 17:10.550 So let's run this first and we'll run this piece of code so you can see for filtered out breathing values, 17:10.580 --> 17:11.320 that's what I do. 17:11.690 --> 17:17.540 Because of this particular situation, the current values for my rating has been converted to zero. 17:17.910 --> 17:25.580 Now, what we do is we will simply open the ratings for value and replace it by minus one. 17:27.300 --> 17:28.710 And when we open this. 17:31.810 --> 17:39.520 Now, you can see the values have been updated now in the trading volume, we have these values and 17:39.520 --> 17:42.790 for those respective ratings, gold has been updated. 17:44.460 --> 17:51.840 Now, we could have done this in several other ways, but we just wanted to introduce to you now let 17:51.840 --> 17:53.940 me get my head again. 17:55.820 --> 17:59.540 Now, in the state of Maine, we have corrected the each column. 18:00.450 --> 18:07.370 The ratings have been updated, the eight show has been updated based on the new age. 18:07.950 --> 18:12.630 Now, the next thing which we can work on is the equal value here. 18:12.630 --> 18:19.860 The fecal value has a string and each string has numbers with a hyphen improvement. 18:21.200 --> 18:28.310 So what we can do is we can split these values now to split a string, we have a split. 18:29.680 --> 18:35.440 So let's try to use that mindset of equal split with hyphen as the split. 18:36.280 --> 18:44.720 So when we run this big legacy, this object has no attribute split, which means that this cannot split 18:44.890 --> 18:47.840 because it is considering this as a CBDs object. 18:48.160 --> 18:52.810 So we need to focus on what this industry and then do the split. 18:53.810 --> 18:59.380 So here they are, converting it into a string first and then performing the split on top of it. 19:01.510 --> 19:05.800 So once we get a list of the numbers. 19:06.930 --> 19:14.910 And now these two lists would have to be converted into data frames, so what we can do is we have expand 19:14.970 --> 19:22.420 method present in place, which allows us to convert this these two lists into a complete form. 19:22.770 --> 19:26.310 So what we can do is if you convert it into a data offering form. 19:26.320 --> 19:32.340 So let me switch this and expand this so you can see that a new data frame has been created with the 19:32.340 --> 19:34.620 forced index. 19:34.620 --> 19:39.720 Has the value from the list the most value from the list, and the second one has the second value from 19:39.720 --> 19:40.110 the list. 19:42.570 --> 19:49.350 Now, the next thing what we can do is here we have got the data frame so we can finally get the Dufrene 19:49.350 --> 19:57.900 convert this into a floor fight because this is still a string like it has been split into two different 19:57.900 --> 20:00.840 values, but this value is still a string. 20:01.080 --> 20:06.240 So we have to convert this into a number, into a plot point and then use it for the. 20:07.370 --> 20:15.260 So what we do here is we get the mighty road vehicle then converted into a string after converting it 20:15.260 --> 20:15.990 into a string. 20:16.010 --> 20:23.570 We split it on the basis of the hyphen and we expand this so that it can words into our data frame of 20:23.570 --> 20:26.390 the string as it is for the string. 20:26.420 --> 20:28.730 So we have to convert it into a floating point. 20:28.940 --> 20:32.110 So we are doing well as they float on top of it. 20:33.050 --> 20:40.670 I'm then getting it into the gate, so this is my new data frame, which I have read. 20:42.170 --> 20:49.130 So let me add on this, so let me bring for you so let me get keynoted. 20:53.120 --> 21:00.800 So she has these two values now what I'm doing is I'm simply creating a new column, if one F2, you 21:00.800 --> 21:05.560 can create one vehicle to vehicle, main vehicle, Max, anything like that. 21:05.870 --> 21:11.120 I simply get the values of key index zero and gain one into it. 21:12.910 --> 21:19.690 So after that, I will simply delete the column because I have selected the vehicle for Elamine and 21:19.730 --> 21:24.280 I don't need this particular column now, I don't want this kind of values. 21:24.280 --> 21:28.080 I can just use the new vehicle values which I have created. 21:28.390 --> 21:31.590 So I'm just deleting it using beneath my dad or vehicle. 21:31.600 --> 21:40.600 I can use different methods for deletion, like throw up my little vehicle so that we can save my data, 21:40.930 --> 21:45.250 not my date of drop, so we can do something like that. 21:46.360 --> 21:48.730 So now let me bring my data. 21:50.230 --> 21:53.740 So now we don't have the full volume and it has been removed. 21:54.140 --> 21:57.700 Now there is a dating column which also needs to be removed. 21:58.760 --> 22:03.590 Apart from that, we have this city in which we need to take care of. 22:04.430 --> 22:06.150 So how do we do that? 22:06.170 --> 22:11.680 We can use the antidote barometer, which we simply need where we want to check the value. 22:11.930 --> 22:17.600 So we are checking my data city values and we are checking if my editors be in the movie. 22:18.170 --> 22:23.370 So whatever might be the equivalent to city might be the city's equivalent of Mumbai. 22:23.690 --> 22:27.350 We are putting one and there it is not equal to my rate as a vehicle. 22:27.860 --> 22:29.350 Then we are putting zero there. 22:30.050 --> 22:38.510 So that will make the City Mumbai column where the values will be one or zero based on if the value 22:38.510 --> 22:40.540 is Mumbai in the city column. 22:41.660 --> 22:42.650 So I just ran this. 22:43.040 --> 22:50.060 Similarly, I have another condition where I'm creating the new column city today and I'm checking my 22:50.060 --> 22:52.110 status to be is equivalent to Jinney. 22:52.370 --> 22:58.690 So wherever this condition will be true, the value will be up in deejayed. 22:58.700 --> 23:04.820 I will be the value will be of not zero and wherever the value will be. 23:06.020 --> 23:10.240 So basically, it will give a boolean exaggerated, will have either true or false value. 23:10.460 --> 23:11.540 They've done this for you. 23:24.630 --> 23:30.480 So when we are checking this, it is giving a true and false value and what we are doing is we are doing 23:30.480 --> 23:31.640 the stip in. 23:32.250 --> 23:38.370 So when the truth will be converted to it, it will be converted to one and then it will all be false, 23:38.370 --> 23:40.820 will be converted to and it will be converted to zero. 23:40.950 --> 23:44.610 So automatically we will get the same thing which we are achieving from this. 23:45.300 --> 23:47.400 So it will give just the same thing. 23:47.470 --> 23:48.660 So let me run this again. 23:49.350 --> 23:55.200 And now we are simply checking another condition in gymnasia global indigency. 23:55.200 --> 23:56.960 So we are just doing the same thing. 23:58.680 --> 24:02.310 Now, our next thing is my little Golgotha. 24:03.270 --> 24:07.800 So we are creating another column called Golgotha and then we are using an order. 24:07.840 --> 24:13.200 So let me just run this and we will need the my column and we see the head. 24:13.380 --> 24:20.250 So now what has happened is the city column has been removed and there the city column value was changing. 24:20.330 --> 24:23.850 It has one here where the city column value was Malai. 24:23.850 --> 24:28.980 It will have one death whenever the city column value will be equal, that it will have one death. 24:29.160 --> 24:35.100 And when the city column value was Delli, it will be all zeros, which is signified by the fourth one. 24:37.150 --> 24:40.150 Now, the next thing is of the gun. 24:41.020 --> 24:44.670 Another thing is we can believe the city, the rating column. 24:44.980 --> 24:49.110 Now here what we are doing is we are getting another dummies for the reading columns. 24:49.120 --> 24:57.610 All we are doing is we are simply doing the deed on dummies and we are taking my data reading and we 24:57.610 --> 24:59.080 are doing a drop forced. 24:59.980 --> 25:06.420 Through, which means that I want to remove one of the values I'm going through, you fix reading, 25:06.430 --> 25:08.990 I'm getting up and you fix them, so I'll just run this. 25:09.310 --> 25:16.060 So here what it has done is we have three, four ratings and one was good, one was bad, one was pathetic. 25:16.060 --> 25:16.790 One was excellent. 25:17.050 --> 25:26.320 So when I used drop first, so it dropped the vibrating column because as we have discussed, it will 25:26.320 --> 25:31.960 mean like if all three volumes of these are Zeitels, that will mean that we are pointing towards bag. 25:32.170 --> 25:34.890 So we don't really need to have the bad column here. 25:35.230 --> 25:39.870 And the prefix simply means that I want this rating as a prefix here. 25:40.210 --> 25:43.240 So in case I don't want that. 25:44.130 --> 25:47.580 And I'll run this again so we will get the bad. 25:48.450 --> 25:54.150 So it has just dropped the first column, so I'll just put it back again and you can see it has been 25:54.150 --> 25:54.530 removed. 25:56.360 --> 26:02.810 So then after this, we once we have created the dummy columns, we can concatenate the dummy columns 26:03.050 --> 26:04.710 and we can remove the ratings. 26:04.730 --> 26:06.740 And here is their data, which we have. 26:07.040 --> 26:11.200 So this is how we can work, whether we can create these rating columns also. 26:11.420 --> 26:14.300 And also we can create this rating scores. 26:14.600 --> 26:22.850 Or the preferred method is to create a rating score, which will create one column only and not multiple 26:22.850 --> 26:24.400 columns for an ordinary data. 26:24.650 --> 26:34.010 This is just to show how Dumi is used, but we will be using this particular method for any ordinal 26:34.010 --> 26:39.500 variable and we will be creating dummies for any categorical. 26:41.780 --> 26:47.000 So this is for this particular session, in the next session, we will learn about the election and 26:47.000 --> 26:53.870 the four steps which we follow while we are creating the features and why we are selecting the features.