Create meaningful fake tidy datasets in R using fakir [#rstats Package]

Key Takeaways

hey welcome to one little coder in this video we are going to learn about a new our package that helps us in creating fake data set and the package name is called 4k so for Karis a package that helps you create fake data set that you could use to teach a better data science or probably you could use to write blog articles or to create tutorials much more so first of all why do we need for kit right so we have got a lot of data sets like iris empty cars which which are definitely used used use and being overused so why do we actually need for Kira's primarily because those data sets are quite you know quite irrelevant for someone who is starting with data sets today so I'm not saying that those data sets are not useful those data sets have been the product of a really hard work of a lot of people so I have got huge respect for word but at the same time when you look at the contemporary data science market a data analysis market so you need data sets that resemble probably close to somewhat industry so maybe you cannot get on exactly as same as industry data set a lot of open data sets are available but when you want to teach you don't want a lot of heavyweight assets so you want something small light that is that is quite you know resembling industry data set and that is where faqir is actually helpful because faqir has a little set of data sets that that resemble industry data set and also it follows steady principle which means you don't have to do a lot of data pre-processing before you when you learn how to do a good amount of data sense and also it has got some product missing values in both categorical and continuous variables so if you want to teach missing value imputation type of concepts also you can still use it and again the other thing is you've got some values variables which are correlated which means if you want to introduce some statistical concepts or if you want to talk about correlation or if you want to draw a scatter plot and then show them how you can fit a line so you can see you some data sets from so let's go ahead and then first install faqir you can install faqir from github which is copy this URL sorry at least the repository's name you can use remotes or you can use these tools it's up to you so I'm going to use three modes which is light install underscore github and say think are open and focus rehab while the installation is going to happen so for Kira's a product of a group called a think hard which is a it is an our development house based out of France they've got a lot of open source contribution for our and this particular package has been developed by Colin fair and Sebastian so a huge shout out to Colin fee and Sebastian for making this package available for us so because this is a new development group from France they have also made some datasets in fakira that is in both French and English there are some datasets that are only in English the sorry French and then there are some data sets that are available in both French and English so what is it asking it's asking do we need to update so given that time so we will say we don't want to update anything just install this package you can see what kind of data sets are available in faqir want a support ticket so if you are familiar with an enterprise business or a consumer business any time you get in touch with your customer support service so that entry gets gets into something called customer ticket right so there are a lot of companies that does this kind of stuff so that this customer ticket you have got what is the customer what kind of problem they are faced and all those things so the other thing that they have got this website visit which is to emulate something like a Google officer term or if you have got a website on it may not be as compliments comprehensive as Google Analytics or Adobe and what it is try to do indicator like the pages that you have got home about blog and contact page and what kind of visits they've got for a given time stone and then they have got some French dataset which is about transport related and then you can actually see what kind of things that they have done with rate asset so they have got quite a comprehensive documentation I would recommend all of you to check it out and then I would also link this in the description page on we can see for Kyra's main sucks fully installed let's say library you okay and you lower the dataset the related something in order so you can see what kind of functions are there for K base clients so you can see the documentation of the the suggestion that pops up that whether it is available only for English or French so this this dataset based client is available for both English and French so Andres is available only for French so on rich people is available only for French and ticket land is available for both English and French or fake visit is also available for both English and French fake wizard Oh some or both English and French people think they have got an issue with the description cool so let's start with fake wizard fake research I will just save assets okay okay listen so now we have created a dataset so let's see what are the parameters that are on available in this fake message so you can see there is an argument that is from date to date and then the seed value which is to say if you want to reproduce this thing you don't want to you know like something random right so you want it to be reproducible so looks like when you look at the data so they've got a one-year or data and so there is a timestamp which is a time variable it's glimpse it okay I've got a load laughter really tidy wash I'll just do it here library tidy verse okay and what do you want to do we want to do glimpse of this it's okay that's that's a type of for me Glenn you can see first one is date second one is a double double then everything else is integer so you can also see there are some initial let's let's count how many missing values we have got let's do a summary the search okay so put a lot of missing values so there the thing is let's say now you want to teach a simple media process to someone so the first thing that you're going to do is you're going to fix missing values right so you can probably say okay visits filled the sets pipe and order what are you going to get you're going to get filled off um everything okay don't you do that thing let's do some somebody who sits fill so you you don't see any missing values here I'd say you've got just one but he wasn't missing while you is filled so now we can probably look at the filled dataset and then we can try to understand okay I think it is miss a filled with and download direction so you can say direction you will not get into that high it's probably not required for this thing so the point is that you have got a dataset and you have got a bunch of pages so let's let's um if you want to teach someone how to visualize something you can say okay yes it's filled plus giome line of AES what is your first value of Israel that is time stamp and then maybe your blog contact so let's say block cool so no no you have successfully made a plot that reduces fuck-you so the point again is from it's just a fake gate acid but you can see that it slightly resembles like a proper time series plot rather than you know anything that is very randomly that you generate from a normal distribution so let's have a look at to the other data set let's say ticket fork a fake ticket client okay so you need to give volume volume let's say in this case we want 100 a thousand it's a thousand okay let's also have a look at what kind of parameters it has got so it has got a volume which is to say the number of tickets you want and the number of claims or other values that you can look at so now let's have a look at this dataset it says that you have got a reference number of clients and you have got the first name and last name of the person and what is their job what is their age okay and you have got ID'd upon let's let's just do a quick summary okay I think you might have noticed that I just simply use pipe operator from the data so that it does generate so you don't have any problem with that so you've got bunch of missing values which we are not going to deal with this moment so if word priority timestamp entry Department some categorical and so there's a mix of categorical and the continuous variable right so that's that's good to know if you if you probably want to do some kind of insight from this thing probably you can start with let us say a ticket I want to look at the first name I want to look at the last name how many times it has appeared right let's its do a ticket count last last name how was it last last yeah it says last and range descending okay so these are the last names AKO call Tom bunch of yes so if you want to make a plot yeah yeah oh sorry sure Jim Melissa Jim bar sorry yes of okay I'll just take the powerhead of 10 or 20 and last comma n okay your score in this case that's an order okay I'll just do in X is equal to Y is equal to the last you have caught a plot that looks designed okay so this is idea the idea of this video is to introduce you to a that you can use for care the data sets from you to teach data data signs especially if you're going to teach on it handling data manipulation data visualization using R and probably you are part of the sky diverse universe it's it's it's it's a good thing to use some something like rather than iris so I hope this video was so useful and if you have any comments please let me know and if you have got any suggestions also please let me know you can also mention me on Twitter if you have any comments on this video so thank you for watching see you in the next video