Julia Vs R Vs Python Programming Language For Data Science

Key Takeaways

[Music] hello my name is Krishna and welcome to my youtube channel so guys from past three weeks have been exploring giulia programming language as many of you had actually the question and told me to create a video on the same atom would Clement at some of the data science project some of the machine learning causes deep learning projects with the help of julia so in this particular video i'll be going ahead with respect to making you understand the basic difference between julia Python programming language and our programming language and you know and definitely I'll also be telling you like how is the future growth if you become a Julia developer specifically with respect to data science enough so I haven't wrote it down all the points in this particular diary my three weeks effort is basically jotted down in three pages in this specific daddy so I will be reading them point by point I have written it in a better way so first of all as you all know guys why python is used it is a general-purpose programming language right if I consider R you know it is basically used for that a data analysis it is used for statistical analysis and data visualization right but when I consider with respect to Julia it is specifically used for scientific computing so this is the first point the second point when was julia created it was created some way around 2012 you know and it became pretty much famous pretty much it just caught the eyes of all the developers in 2018 and why because i'm just going to say you some of the points as we go ahead with respect to that now the third point that that the quality that julia has you know julia is also a programming pretty much similar to the Python programming language you know you have beautiful syntaxes you have functions oops concepts many more things you know with respect to that now what is the additional benefit that julia has is basically the speed and performance and this speed adds performance are almost similar to the c programming language right in c programming language you know that how the speed and performance is so similarly that particular property julia programming language is actually having so it has literally caught the eyes of the developer community so coming to the next point based on the community that the people are actually working in fightin fightin fightin and har has a huge community fest it is a huge huge company community because that you you know that it is one of the loved programming language one of them wanted programming language right whereas if we consider the julia community right it is very very small yet it is growing up probably it may take around three to five years to go into a loss scale as such as Python now why this community is pretty much important because I understand whenever you're implementing some projects with respect to Python by using some wonderful libraries like tensorflow Chara's and if you're stuck in some place right you can go to stappord so you can go to github and you can actually you know ask that particular covariance someone will be there to reply it and why that is possible because you have a huge community were working in this particular domain in this particular programming language itself right whereas in Julia the community is pretty much smaller but yes the developers are very very friendly you can you can see that there are lot of tutorials that have put up in YouTube channel if you go and put up your queries in Stack Overflow also you'll be able to get answers but again as the community is huge Julia and Python if I consider Python it takes that particular place right so coming to the next point about the libraries now if I take an example of Python Python has more than 200 K libraries guys ok 200 K libraries so that is pretty huge if I consider our programming language it has somewhere around more than 15,000 libraries but if I consider about Julia it just has somewhere around 3,000 library itself so if I take in this particular example definitely again Julia is having less number of libraries when compared to Python you can also say libraries or packages that is basically used for doing statistical analysis for doing exploratory data analysis you know for doing machine learning algorithms so those kind of libraries if you're combining it has some way around 3,000 libraries itself ok now coming to the next thing is that how many number of downloads have happened till 2020 at the month of May ok so that also I have noted down it is having more than 13 million downloads okay Julia so definitely buy this number of downloads you can see that the developer community has really got an eye on Julia and this is all because of the speed and performance you know you can do almost all the lifecycle for data science project also with the help of Julia because I have done it literally if you want some tutorials please do put down your comments in the comment box I'll definitely be uploading tutorials as we go ahead okay now the next point that is to be noted with respect to julie and i can say that this is a kind of drawback there is a concept which is called as time to first plot problem now what happens in julia is that where as soon as you use a package you know it is just a kind of just-in-time compiler so that basically means as soon as you load a package it takes some time for compiling at the first time you go because it has a property of just-in-time compiler but after that when the session is open for a longer period of time then you can actually speed up the process okay so this is just a drawback with respect to this and the biggest selling point with respect to Julia you know because that is the most biggest advantage in Julia itself that is speed and performance okay so there is wonderful examples that has been given why Julia's programming language speed is very very high it actually comes towards the second position after C programming language okay there was a recent research that was done where you know or a developer what he did is that he tried to take more than he tried to generate more than ten million random numbers and then he did a quicksort on those particular numbers right ten million random numbers so when he actually did it with Python programming language it took us it took him to actually take somewhere around 124 seconds whereas with the help of junior programming language it took somewhere around 24 seconds but this is with respect to 10 million random integers okay this is a huge number guys so if we also consider guys there are some graphs in Wikipedia also which says that Python Giulia is better than forego programming language photon programming language Java JavaScript she shop peyten our programming language with respect to speed and performance but if you consider just a small number of integers or small number of values with respect to that same quicksort almost Python and Julia are actually giving the same amount of time so that basically means that Julia works well when your data is huge you know when the numbers are huge right and considering data science where you will be having huge amount of big data at that time even a microseconds can actually be very very beneficial for the developers yeah so this was the next point now coming with respect to the lifecycle of a data science project now guys as you all know that we have various lifecycle of a data science project like you know we have to do the data analysis exploratory data analysis we have to do the feature engineering feature selection machine learning clear model creation then you have to do the high profile entertaining of the model and then finally you have to do the deployment now considering this guys with as you know that in power in Python you know for doing most of the exploratory data analysis feature and getting we have a library called as pandas similarly in if I talk about julia programming language there is a package use it called as data frames got GL okay so this particular the package actually works as if like how we are actually working with pandas in python okay you have all the group by function you can concatenate it you can drop you can drop columns you can do all the functionalities like how we used to do in python then there is also with respect to visualization libraries like how we have matplotlib got pipe lock similarly here we have something called as plots package okay and there are some more advanced packages have not yet explored but I have explored plots package where you will be able to do lot of data exploratory data analysis with respect to various kind of works that are present then so in Python you know various integrated development environments like spider you have pycharm with respect to our programming language you have our studio right in case of Julia you have a wonderful ID which is called as Juno Juno is almost similar to a art studio I've I've seen this and there is also a package which is called as I Julia which will actually help you to integrate your Julia programming language within your Jupiter notebook that you are using for Python so I had actually installed this particular packages it called as I Julia and I try to integrate with my Jupiter notebook and I did the programming over there and probably after Sunday's I will also be coming up to this particular tutorial so that you'll be able to see it the next thing is that with respect to deep learning you know for deep learning also julia has wonderful packages there two packages specifically I would like to say is one is flexed or GL garage door jail and yes I missed out with respect to machine learning packages in machine learning packages also you have packages like sky Kaitlyn George Al you have for statistical analysis you have stats model dot GL so these all are like kind of packages name you don't need to import it you need to install it and then only you will be able to use it this prop this packages also has a property like just-in-time compiler okay you have to first of all compile it then you'll be able to use it and make sure that you have a session open for a longer period of time and that I'll be showing you in the upcoming videos as I'll be developing about this particular julia programming language with respect to data science now as you seen most of the properties most of the things Julia is having when compared to Python and it also has an additional quality of speed and performance but they are something like you have less number of packages the community size is yet very small when compared to Python and many people who are actually doing research they also deserve doing research with considering Python programming languages and open source now the question comes is that whether you should learn it or not I would give a very good idea to you guys start with Python programming language or our programming language in learning data science you know what I would suggest is that don't directly jump to Julia you know get the understanding is that how you can actually implement Python programming language with respect to the data science functionalities you know then once you are able to get a very good hands-on you know and then when you are learning Julia right that will be actually beneficial to you why because understand in Python the community is very huge any type of issues that you face you will be able to solve it right in Julia the community size is yet small it is growing you know it will be growing and and the probability at time that will take you know to become a full-fledged like Python or R it may be somewhere around three to five years so what based on my experience I would suggest that start with Python program language and yes in the future you can start exploring Julie a programming language also like I like how I am actually doing a la I've also started exploring Julia programming language but I have actually gained some kind of expertise with respect to Python machine learning deep learning techniques specifically when I am implementing with the help of Python programming language and slowly slowly now I'll be also moving to this to understand at how fast how quick what is the speed and performance when I'm actually using julia programming language so with respect to the future growth guys probability will be taking somewhere on three to four five three to five years you know for the julia to become you know accumulated or used by so it may probably take three to five years for many developers to use julia and there is one more disadvantage that are actually final dot is that there are some bugs you know with respect to some of the packages that are used in julie also so this was one of the thing that i'll actually found out and yes i've covered pretty much all the points but my suggestion is that first of all start with python programming language or a heart programming language learn some kind of databases try to learn machine learning and deep learning algorithms and then try exploring julia because once you get a hands-on experience with respect to python learning julia again is like a cup of tea you know and you'll love it because just understand within three weeks i was able to explore machine learning algorithms in a proper way i was able to understand how it actually works i was able to implement a deep learning algorithm I still have to explore that how we can use GPUs with respect to julia packages that we have libraries so that i need to understand and i will probably take some more time but yes I had just invested somewhere around half and half hour or one hour every day and I was able to cover up so much so yes it is also possible by you guys so yes this was all about this particular video I hope you liked it please do subscribe the channel in order to subscribe it's all in the next video have a great day thank you one and all