Applied Data Science With Python Full Course 2026 [Free] | Python For Data Science | Simplilearn

Key Takeaways

Hey there, welcome to applied data science with Python by SimplyLearn. Let's start with a quick question. Have you ever wondered how companies predict what products you will like or how search engines gives you exactly what you're looking for? That's all data science at work. And guess what? Python is the tool that makes it all happen. In this course, we are going to explore how Python helps us handle data, make sense of it, and even turn it into something useful. We'll focus on three key tools. Numpy, pandas, and mattplot lip. Don't worry if these sounds new to you. We'll break them down step by step. Here's what we will cover in today's video. First, we'll dive into numpy where you'll learn how to create and manipulate arrays. You'll also cover how to do all sorts of calculations like finding averages, medians, and more. Basically, making your data do whatever you want it to do. Next, you'll also get handson with pandas. You'll discover how to clean up messy data, handle missing values, and transform data so it's ready for analysis. You'll also talk about turning text into numbers, which is super important when you're working with data. Then, we'll talk about mattplot lib. This is where the fun happens. You'll learn how to turn all the data into cool visualizations like charts and graphs. You'll see how visualizing data can help make better decisions. Along the way, you'll also talk about the real world data science process, how to figure out what problem you're solving, how to explore and understand your data, and how to use the right tools to get results. By the end of this course, you'll be comfortable using Python to handle data, clean it, analyze it, and even show it off with visuals. Whether you're just starting or looking to sharpen your skills, we have got you all covered. So, let's get started to unlock the power of data with Python. Before we commence, if you're interested in mastering the future of technology, the professional certificate course in generative AI and machine learning is the perfect opportunity for you. Offered in collaboration with E and ICT Academy, IIT Kpur, this 11month live and interactive program provides hands-on expertise in cuttingedge areas like generative AI, machine learning and tools such as chat, GPT, DALE and also hugging face. You'll gain practical experience through 15 plus projects, integrated labs, and live master classes delivered by esteemed IIT Kpur faculty. Alongside earning a prestigious certificate from IIT Kpur, you'll receive a official Microsoft badges for Azure AI courses and career support through Simply Learn's job assist program. Hurry up and enroll now. Find the course in the description box below and in the pin comments. Here's a simple quiz question for you. What does pandas help you do in data science? Is it to visualize data, clean and organize data, create machine learning models or is it to build websites? Let us know your answers in the comment section below. >> Next, we'll talk about the concepts about the data science part. What is data science? What is the process and so on. Again, that's the first page right that I have as a part of think. So a fundamental definition that we goes through whether there is use of scientific methods, processes and algorithms to derive meaningful insights from the data. That is the idea behind it. I will not go through with that right. I will not talk about this. What I'll try to talk about is to think of a real world scenario. I'm giving you guys a scenario. Can you on everybody I hope so everybody can see my screen. on my screen. Can you see there is a bulb? Yes or no? Great. Thank you, Gilbert. Thank you, Priti. Now, that's data for me. I'm giving you this data. This is a data. It's a visual data, right? It's a visual data. Now, think yourself as a person who is a data scientist, right? who is a data scientist at some bulb manufacturing company, right? Some bulb manufacturing organization, right? You're a data center as a bulb manufacturing organization. You're walking out of your house and you saw this bulb. You're walking out of your house going to work. Imagine you're going to work and this bulb was there into the next house or maybe somewhere on the road you found out this particular or somewhere on the pathway you found out this particular bulb there. Great hypothetical situation but thinking process this is the most important part for a data scientist for an AI engineer is the most important part is the thinking process. How do we think about it? Now think about what you can recommend as data scientists considering this particular bulb as an input right this is the input to you think about that okay you may not know what is data scientist roles but whatever your understanding of data scientist based upon that right or AI engineer course right whatever you course enrolled into what is your approach approach towards a business proposition or about this particular information that you have that you can recommend to your let's say the uh to the management over there maybe you can tell okay I saw that bulb on the street and I have so many so so and so recommendations for you anything that you would like to talk about or let's talk about what you can tell me about this bulb let's say how will you draw insightful information about the data. Can you tell me about this bulb? Just information. No decision making. Just about the bulb. Tell me what all information you can give me. I'm okay with that. You say that if the bulb is yellow, that is also an information. It's a screw based bulb, not a click. unlock method. Uh, it's brightness level. Okay. Bulb dimensions. Great. What else? See, I'm looking forward to words something. Power consumptions, great. Materials used, wattage, great. Lights when connected power, okay, that should be, you know, a basic property. PRI. Yeah. But it's still still good. Anything else? See, I want you guys to observe thing. Take your time. Observe because whenever you are saying, let me just try to tell you. I have too many options came up like Abinitri says brightness level is it observable from the given image. The Bible dimension Gilbert is it observable? Is it available on the given data? The power consumption. Narendra says that wattage or the material used lights when corrected power. How many watts is that available with the given data? Can you actually imp you know find it out from the given information? Uh Ajay Shri what is that? Yeah. Yeah. Sam is coming up to caution validation required before green. on top some black dot is there. Yep. What does that implement? E27 bulb holder. Agreed. Uh Dashish. It is an object. It is a object. Okay. Any object. Good. So I tend to you know closely follow the observation from Sam. The kind of inference that you might require to take it up is something which is unusual right something which is unusual because as a data scientist or as an AI engineer uh everybody knows right you know I know you can see the options coming from 10 different people's in the chat window as well someone is talking about you know wattage powers luminosity and so on so forth Right? But something different, something that makes you unique, something that is no one else is able to observe and that has some meaning. So sam talked about there is a black dot on the top. What what do you recommend? What is the you know reason behind it or what would be the implication? Uh it's a filament bulb. Great. It's a filament used in the bulb. It's a tungsten bulb. Great. So one observation which is you know directly not available but it is hidden. None of the bird manufacturer label that's not available right the bird but what I can observe let me just put it across what I want to convey to you guys is right with this particular image see here I can see that there is a black dot on the top right it's a black dot on the top that black dot tells me that whatever the outer covering is right of this bulb that outer covering is getting blackened out that means the intensity of the power inside the bulb is such high that it is turning out the features outside or the outer body is black. And if that is happening that means it has to be usually a tungsten bulb. That's number one. It's a tungsten bulb. Second thing, this does not happen over a day. This does not happen over a day. It happens over a period long period of time. It also tells me it's a very old bulb, isn't it? Do we agree? Thank you. And if it is a very old bulb as business perspective, I can think of Yeah. incandescent bulbs for example. Yeah. Similarly, it's a very old bulb. Now, I can think of an alternative bulb. I can think of as a business proposition. I can think of an alternative bulb that does not give the black structure gives me better you know life or maybe something like that it's not using tungsten technology maybe using some other technology like light emitting diodes could have been any other solutions right which will offer a better or cheaper options make sense now why did I discuss about this bulbs for so wrong. The only reason I wanted to talk about is not the bulb but the approach. Make sense? See, I can go on with this theoretical presentation for a long period. Not a problem. But the point is this is not what we here to learn. You can go through with this presentation 10 number of times. You will have access to this presentation. So I will not concentrate more on to the presentation part. what we need to do in real world. I'll try to concentrate more on that. Is that okay with everyone? Right. Presentation is something that you guys can read through 10 number of times. Right. And most of the slides are self-explanatory on this. So what did he learn from here that whenever you're getting a data whenever you're going to work around with a data you do not look for things which are obvious right you need to look for things which are not obvious isn't it now in order to find out those things you require certain kind of methodology here it was easy for us right because we can see that pictorially there's a picture available and I can see that part I You see that shine there is a you know slight little curvature onto the active I can use that too for some informative purpose like it's in a background where there is too much light that's why I have that shiny part or shiny curve on the top that means there's a background light reflecting on that particular bul right so understand this part that it is not about that what is given already in the data it's not important because everybody can see that what is important is what is hidden in the data is that okay thank you hopefully thank delivered. Okay. So moving on now I have some you know basic things to talk about very quickly walk you through with that. So the process of data science it's a kind of you know process where I would require lot of information right I need to know about the domain I need to know about the scientific methods and technologies very important things right now domain is something that you may or may not be aware about at any point of time for example let's say if I talk about when I started my career as a data scientist I was coming from a financial background I was working state bank of India prior to So I came up from a BFSI background banking and financial institutions. So my core expertise was on finance but if you talk about over the period of time what kind of projects that I've dealt with I have dealt with the financial non-financial pharmaceutical I've aviation different kinds of projects that I've dealt with. So subject expertise is not something that you will have in all sections ever. This is something that you will acquire project twice. Right? You would have to do the hard work to read about the project, read about the domain, understand the domain, acquire knowledge about the domain. That is what you need to do, right? When it comes to the scientific methodologies, right, that you need to do. So scientific methodologies are the mathematical tools and the scientific tools that we have, right? So I'll talk about that during this course. What about the scientific methodologies that you need to require to be a data scientist or any engineer? We'll talk about those mathematical and scientific concepts that are you know used in the real world. We'll talk about that during this course. As a part of technology, we require a machine and a programming language. Two things, right? A machine and a programming language to implement these things. Good. So I have a programming language like Python. I can use that. There are other languages as well which allow you to do that. Then operating systems, data processing tools. Again, we'll talk about the Python language slightly, data processing tools, the libraries, the application design. We'll talk about all those things during this particular course. So that's your upcoming portion. Are we good? Keep on asking questions. Some of the applications of you know data science processes is being given up here. One of the data science process application is the biometric sensing of the data right you wear up a you know wristwatch or the you know what you call it as a smartwatch it counts up your steps it counts up your heartbeats and talks about the real-time information about that I can use that information and take out some decision- making process and make some contributive decision let's say suppose my my wrist or my this particular smartwatch detects that my blood pressure is going beyond 200 right my blood pressure is going beyond 200 there is no movement in my body can be detected now might be possibility that I'm suffering from a cardiac arrest or I'm suffering from you know hyper blood pressure so I might require immediate you know emergency services help uh my Android device or my um you know whatever Apple device or whatsoever it may be right My smart device can make up an emergency call to the emergency services and send my location. If it can send my location, emergency service can arrive at me and take necessary care, right? Without any human intervention, we can still deal with these things, right? So useful interactions that can happen. Another example of data science is the search recommendations. So recommendation systems we yep not in this course but in the next course machine learning we'll talk about that. How does the recommendation system feature works? We'll talk about that. But yeah, recommendations that you might have observed on Google platform. You write down typing something and it starts you know showing up. Okay, this is what people have searched for or this is what is the possible next word. So next word recommendation right suppose I want to write down um maybe you might have seen in how many have seen that okay while you're writing a mail right it tells you if you write on hello it automatically shows you sir or madam and so on it asks you to press tap to select that option automatic recommendations of certain words completing a line completing a paragraph seen that Therefore expecting yes. Yeah. Thank you. So that is coming not from just you know any kind of software. It's just a data science process. The recommendation systems. Similarly your loan applications right? Credit card applications or loan applications. They say that you'll get an approval in 2 minutes. Who's going to look after that application in 2 minutes? No one. It's a backend machine. Looks at your credit report. Looks at your history. looks at you know your application studies and everything and makes up an informed decision based upon that data right you'll study about how does the decision-m process happens in ML course machine learning course right you'll study about that but as of now you can understand that this is an uh you know an application of data science there can be a lot of applications right uh any field that you think of there is an application any particular field you want to put up an example any specific field you think that I should put up the example with I can put that example no problems but ask it any specific field for example let's say you apply on on a job portal for a job you think that my job that my qualifications are exactly what is been required for a particular job you still do not get a Thank you very much Abinitri. See this was you know theoretical portion so I was not expecting too many questions with this part so far but I just want to follow the process. So this is the entire process that we have right in the field of data science right now in this there is few things right the first thing is problem definition. Problem definition is a kind of a problem statement that is given to us right a client will come to your organization or to you to tell you that okay this is the particular problem that I'm facing particular business right say for example I'm a soap manufacturer I manufacture 10 different kinds of soaps out of those only one particular soap is not getting a high sales or the sales values are very low for that why is that happening no you need to answer as a data scientist or an AI engineer what what is the problem what is the possible cause of the problem and how we can rectify that problem right so once we know the problem based upon that problem we ask the client to provide the data right the data acquisition happens now what is the skill set that is required for data collection anybody I'm I'll only talk about the skill set as a part of data collection. Usually what happens is a client will tell you okay these are my you know database servers I have kept my entire transactions data or whatever the things that we have I have kept in particular database these are the login credentials that you will have take out these login credentials and fetch the data from there sensible so we should be aware about how to fetch data from different database servers now as a data science if you talk about end to end process uh people will seek those kind of things. If you talk about small or marginal organizations, very small organizations, they think of person who can do end to end job. But in bigger organization or mid-level organizations, they have a separate team who does the ETL portion. We call it as extract data and transform and load it back and then we have a different set of data scientists. But in general, the skill set that is commonly expected is structured query language. SQL not a part of this course. We'll not talk about SQL at all. That's a separate course altogether. You may have that as an optional topic but you should know about structured query language if not in detail but slightly about it. Are we good? So that is where we collect the data. Either we'll do that through databases or we'll you know roll out Google forms. You see that there was a Google form rolled out to you at the beginning of the session. Isn't it? What was the purpose behind it? Yeah. There is one purpose that I get to know about you guys. But there is another purpose that there's a data collection process that is happening in the background of simply lab. They get to know you know uh those or whatever the data scientist team that is working in the background for simply learn. They will get to know okay what kind of learners are joining in what they know about what are which industries and so on and so forth. A lot of conclusions can be made out of this. Lot of business decisions can be taken out of that. Are we good? So data collection is not just with the SQL. It can be directly from the you know from the end users. It can be directly through you know online resources. You can extract it from web pages. You can just you know you know walk through with the web pages and you can use that. Maybe there is another library that you can use for data collection is BS4 beautiful soup that is a web scrapping library but that's not our objective. We're not going to talk about these two portions. the problem definition and data collection. Why? Because we are going to go with a standard problem statement. There is a standard problem statement that will be given to you and there is a standard data. Right? Which is in the format of CSV. It is in the comma separated value format. So what I'm expecting is you will have a data set, you will have a problem statement and then we'll start working. Are we good? You will uh get this the problem statement maybe I'll address it live in the session. The data set if it is available into your you know standard you know practice data sets it is available into your LMS otherwise if I will use some additional data sets I will share it with you guys. Okay, if I would require any one of those, so our learning for this particular course will be limited to this part only the part two and part three. We'll get to know how do we clean the data, how do we explore the data and how do we perform the feature engineering part. So this is the basic thing that we'll do here. Right in between there is a lot of things. How do we clean and explore the data? Exploration is the major part, right? Because we need to make a lot of decisions about the data while doing the exploration. Some part of model building is also, you know, part of this particular thing, right? We'll discuss into that part. But we'll not actually build the model. Maybe if time permits, I'll try to see some one or two models to be built up, right? But this part is not part of your curriculum. Model evaluation and model deployment. We will not discuss that at all. This is a part of your machine learning section right next course machine learning these three topics five six seven. So essentially the entire nine sessions are dedicated to this particular portion only. Now how important is that is you can understand this suppose if there is a project let's say there is a project of data science that talks about a particular problem and the time span given is 6 months to complete that project. For example, out of those 6 months, you can imagine at least four month will only be spent on this part. At least four month. It can go up to five as well. This is very very critical factor because how good we clean the data and how good we manipulate the data that makes the next stage which is the model building more robust more efficient. So that is one of the very critical factor in the field of data science or machine learning or artificial intelligence. How is the data being prepared? So data preparation is a really important topic and that is what we're going to learn through this course. Are we good? Okay. Python for data science uh preferred programming language across the industries have some advantages and disadvantages. We'll not talk about that. What are the important packages? Numpai is one of the important package. We'll discuss that in detail. Pandas is the second topic. Sci we'll discuss that shortly for some period. Then stat model will come into the picture as well. We'll talk about scikit learn uh not too much in detail but slightly about scikitlearn as well wherein I'll talk about what is scikit learn and some features of scikit learn will come into the feature engineering course. Matt plotly will discuss in detail. Seabbon will discuss in detail. Plotly will come with the you know questions. Okay. With the plot loops. This is a types of plots with the examples which is given into the presentation. Uh wherein what is a line plot? Connect two dots and make join them with a straight line that becomes a light line plot. A marker plot. Highlight those dots and that becomes a marker plot. So you can see your dot dot dot dot. So that becomes a marker. Then there's a scatter plot uh which is typically you know used for the purpose of analyzing relationship between two data points right we'll discuss all these plots in detail this is just for your reference to know about these words that's it then you have this area plot you know that is also called as a stack plot area plot and stack plot are same thing and it talks about you know cumulative changes between certain datas right we like see why I'm trying to rush through with this is because I need to talk about this how these are created and what kind of meaning they take up at a later stage as well. So I do not want to put up time here just to talk you know if I give you details right now and then later on again same problem arises. So I need to talk about how they are created and how we will use them. We'll do that all these graphs that I'm talking about. Bar plots again talks about you the frequencies of the things how many things are appearing how many times grid plots drawing of the grids histogram another plot very important plot that talks about the frequency distribution pie chart again talks about the frequency distribution that's it what is data exploration will we be diving into the specifics in a subsequent session yes definitely in the data exploration part we will talk about a lot of things pretty there will be a lot of things as a part of data exploration essentially when I'll try to start working with the pandas library right when I start working with data pandas library I'll tell you how to reach out to particular why to reach out to a particular data what inferences you are going to make out from a particular data and how will you deal with them every kind of possible scenarios the different sort of data sets I will try to include that whatever the possible situations that you may come up across in a real world scenario up to a certain limit right because I have limited number of data sets that I can talk about but yes I will talk about data exploration essentially that will be the most longest thing that we will have highest longest thing after the mathematics portion the maximum time we'll go on to mathematics but next thing that will be data exploration otherwise I just need to talk about syntaxes right suppose what is numpy what is numpy syntaxes what is an array so basic concepts right so that will be a syntactical thing where I'll talk about syntax and tell you okay this is how the things are designed in python we need to follow the struct very strict pythonic way that is one thing I'll tell you this is this is what is supposed to be done that is fixed but when and why is something I'll let you get all right so what I will do here is I am currently using Anaconda you know tool which you can do it by this way you can go on to anaconda.org download.success success. I have shown that link and you can download your requisite install installation. If you're using a Mac machine just in case if you're using a Mac machine, you can go on to the Mac link and find out the downloader here either for silicon uh Intel chip is not available there. I am not sure but uh Apple silicon graphical Apple silicon command line is Intel chip is also should be available right so if there is any concern with that let me know otherwise you can get it from this this is a local installation that means that allows you to create files locally on your machine use your local hardware and work on the local mana manner but typically you can also use you know you a a cloud cloud based hardware and that is what you can use with the collab. So you can search for golab.resarch.google.com once you go onto that page that should appear up with something like this if you're already signed into your Google account. If not it will not appear up that way but you can click on the new notebook option and you will come across with this kind of scenario. I uh office laptop is VS code install extension of Jupiter. I am okay with that sites. I'm okay with if you do not have to install that extension and you're putting up a path for Python up quoting uh yes simple does offer a lab you can use that but what will happen is in a later on courses like when you will you start machine learning or deep learning some later courses simply in lab is going to be not available and I do not recommend people doing the simply learn lab. See, I am working for simply learn but I do not recommend it. Why? Because there is a reason behind it. The reason is because some of the libraries in in that particular field are not current. They are slightly older one. They're not updated and hence what happens is certain functionalities will not work in a proper way or they will work but they will not offer you this solution that you should expect out of it. Right. So I do not generally recommend people using simple. That's the the final you know ultimate thing that I tell people. Okay. If you cannot do anything go with the labs. Great. I will recommend you to use collab. Anybody who's not okay with your local installation come to the collab page. That's how you will see it. I will just close out this you know AI companion and I'll put on one simple code print my name very simple thing I write down that piece of code enter my cell block over there and I can execute that by clicking on that play button it gets executed I am not going to use collab in general I will be using the locally installed Jupyter lab environment. So I'll show the same piece of code on the local Jupyter lab. Here it goes. So I open up Jupyter lab and I have clicked on this sheet and I can write down that print same piece of code and I'm putting up my name. This is going to be faster for me, right? It works. Hopefully 1 2 3. It does not. Why? Till now taking too long something is wrong with my machine. Maybe I'm acquiring too much of done. Great. So anybody and everybody give that a try. Can you just take out that print statement and see if you're able to execute that? either in collab either in uh your environment whichever you are preferring I'm okay with that but just make sure that you are able to get the codes executed AB3. So we'll start talking about the first library and that will be your first file which I'm going to rename over there numpy and I'll put up that is for your Jan or rather this is your ADSP batch for January ADSP B1 that's batch one for Jan 26 or January that's fine Great. So let's start talking about numpy. So what is numpy? So, numpy is essentially a framework or library one of the very basic or based library that is available in Python or that's a part of Python which is used for numerical computation primarily used for numerical computation. So, NumPy is a library of Python which is used something from SA. Yes sir, I'm telling which is uh numpy is a library of Python which is used primarily for numerical computation. The word numpy is coming from numerical Python. So numerical uh that's fine numerical Python. So I just highlight those characters in upper case. Right? So essentially the idea behind the numpy was that I wanted to make sure that I can use the mathematical concepts of python in an efficient manner. Right? It brings in what does it brings? It brings the concept of brings in a new data structure. A data structure that is array. Now do we know what is the data structure? Anybody? Not data types. It is a data structure. Now data type and a data structure are slightly different in nature. Narendra collective data of different types not even of array is not of different types. A data sets are not even a data set either. So when I talk about a data structure, a data structure, a data structure actually defines defines the arrangement of data defines the arrangement of data in the memory. Let's try to understand that through a conceptual or graphical manner. Very simple thing but I'll try to put that into a graphical manner. Suppose it's your birthday possible. Anybody's having birthday in the month of January. Oh, January is finished. February in the house. Suppose it's your birthday. Now, there is a locality, right? There's a locality where five of your friends are living. There are five friends of yours. Suppose this is the entry gate of that locality. This is your friend one's house. This is your friend's two's house. Right? This is your friend three house. This is your friend four house. And this is your friend five house. Now the problem is that you know where is friend one. So you can reach out to friend one give him the invite. Friend one knows where friend two lives. Friend two knows where friend three lives. Friend three knows where friend four lives. And friend four knows where friend five lives. Good. Now if you think about this process, it is going to take up some time for you to go around with all the friends houses and distribute the invites. Yes or no? As compared to as compared to when I ask you the comparison if it is the situation like this that you have a friend one you know where is friend one but the friend 2 3 four and five they are living adjacent. You know that friend one lives at a particular place. You will go up to the friend one and then you know next house is of friend two then three then four then five. Now definitely you will take up less time in this scenario isn't it? Thank you. So that is what array offers. So array is a arrangement of data into the computer's memory in a contiguous manner. Contiguous means in a single chunk. So whenever data is arranged, right? So what is an array? That is what numpy offers. an array which is referred as ND array. The word that is used in Python is ND array. Right? We use the word ND array in Python. So referred as ND is a contiguous contiguous memory allocation of the data of same type. Now idea is that every information will be of similar nature and it will be arranged in a contiguous manner. Contiguous means that one after the other it will not be scattered around in the memory but it will be together in single child. Sensible, not sensible, yes. No, maybe you need to ask questions. Now every time when I start with numpy and I say that it is used for numerical computation people say that python has its own module for numerical computations like maps then why do we require numpy so that's the next basic question that is why numpy why numpy so what is the numpy because of the use of the arrays and its functionalities, right? Its functionalities. Numpy is NumPy offers much faster execution that means it is intended to be very very fast that is the only reason I require numpy then the core python let's try to look at something now first of all before I begin with this how many of you know what is an import statement import keyword anybody body have ever used import keyword before? Yes. No. Maybe. Yes, Narish. Yes, great. Yes. Yeah. For importing. Yes. To use the library. Thank you very much. So, almost everybody knows it. Hoping so. Now considering that let I'm going to import few things now in front of you right I'm going to import a library that is called as import ah I'm going to import just correct that it's raw make it code let's import time one of the library which is the default library to Python which is time library I'm going to use that library why I want to use that library is because I want to showcase something. Great. What I want to show is this. If I just talk about time dot time is a library that I'm using dot time and I'll try to print it up. Let's print what is the time dot time function is going to give me. Now it gives me a epoch time. What does it tells me? It prints the epoch time at execution. Right? What is an epoch time? Somebody help me. Present time. No. Unix time. Yes. But what essentially that means? Unix time. Narendra. So epoch time if I talk about is the time that is elapsed in seconds since January the 1st. 10 01 1970 right so that is called as an epoch time right epoch is another word that is used alternatively as you know Unix time or Linux time uh essentially that was the period when Unix was introduced and the clocks were reset to be standardized for you know a standard reference so it was you kind of you know something of that we follow in the computer science fraternity to you know keep a hold of time as a fixed things across everyone in the entire globe because time is something which is varies across the globe you know every geographical location I am currently right now it's almost like 11 uh you know in the morning or early afternoon you can say that 11:00 a.m. is the timeline for me. Some of you would be at different time zones, isn't it? For some of you, it might be early evening, for some of you it might be the early afternoon, maybe late afternoon, some other right, isn't it? Depending on the time zones that we follow. So in order to standardize that we use a particular timeline on the machines and that is we call as the epoch time or Unix time or Linux time. Great. Now what is the advantage over there? Now since I have executed this particular statement right there is some time elapsed isn't it? Let me just try to copy the same statement again. Paste it back again. Good. And execute. Can you notice that there's a difference in the two outcomes? Right? 1769836489 176983 66649 or essentially if I can say that between that two executions right I have there was 159.569 seconds elapsed this is a difference between the execution time Right? Two statements I have used and I have tried to you know showcase the difference between the time of execution. Are we good? Now I'm going to utilize this for what purpose? The purpose is to tell you why numpy. Good. Great. Yes. No. Maybe. Now I'm going to use out a data structure over there just to showcase this part which you will learn about in a due course in upcoming time maybe today itself or tomorrow that is going to come up right but I'll start talking about this part right and that is the a random thing so I'm going to create an array the idea is to let's create an array create an array of random data. Create an array of random data. Right. So how will I do that? Since I am using that, you can use that in uh you know syntax for yourself as of now. Do not put into your brain for this. Just think of that as a syntax. Okay? And I am expecting everyone is going to do that for me. Right? I'm going to create an import library import numpy as numpy. Numpy I'm going to put that as a name np. What essentially means that I'm importing the numpy library with an alias name np. Thank you very much narendra. And now I'm going to create up a data. So let's say data is equal to np dot random dot rand and I'm going to use uh let's say 1 1 2 3 4 5 6 7 8 zeros. So what I want to do it is to ask my numpy to generate a random values of what nature of random nature. Now rand means a random floating point numbers between the range 0 to one. We'll discuss that rand function later on. I'll do that. But as now consider I have a floating point numbers which are you know decimal point notation numbers that will be between the range of 0 to one. Good. I'll have that many numbers. Good. It generates that data. Good. The data is now generated. Now if I ask you that I want to calculate the average of this a very simple thing right there are that many number of data points that's 10 100,000 10,000 100,000 1 million so that is approximately 1 billion good 1 billion numbers if I want to calculate the average of those 1 billion numbers how will I do I need the summation of all those. Agreed. What is an average? Take the sum of all numbers, right? Divide by the count of numbers. Is that okay? That's average. So, let's take a sum. Sum is equal to or let's say I'll define that average is equal to whatever the data that I have. So I'll take up the summation of that sum of AVG that is going to give me the sum of not AVG but uh DATA I'm taking that sum divided by the count of the count is not available the length of DA good alen of DA is going to give you the same data if you just let me just show you that number right same number. So I'll calculate the average over there with this function. Good. Now I'll try to do what? I'll try to note down the time. So I'll take up a start time that is equal to time dot time. So that is going to capture the time that is elapsed since epoch before the execution of this statement. Good. And then I'll copy the same thing and I'll put it down below and I'll put that as a word stop and I'll write down time up time again. So stop is going to capture the time immediately after the execution. Agreed? Thank you. Let's try to display what is the time taken. So can I put a print time taken Z equal to and I'm putting that as the stop time minus the start time correct everybody okay with this till now thank you I'll execute that for you that's going to take some time. Let's see how much time it takes up. It is going to be machine dependent. My machine will take some time. Your machine might take some different time. But it takes me 12.69 seconds. The difference that I got is 12.6994. So approximately 13 seconds. You can say that. Good. Now I'll copy the same piece of code. Right? No changes at all apart from how I am going to calculate the average. I'm changing the process of calculating the average. Instead of using the you know the traditional approach I'm going to use a numpy library approach I will ask np.m mean. uh that's a numpy function to calculate average of a data and I can put up that I want to average of this data which is by the name DATA good and look at the time time taken 0.42 42. Do you notice the difference? Right now, some of you guys, some of you, I'm not expecting all of you, but some of you are going to tell me how much time on your machine it took up without numpy and with numpy. Some of you will tell me that will confirm that everybody is doing along with me or not. This is just for that purpose nothing else because what I'm expecting is that you guys will do along with me. I have not yet started with numpy. I'm still trying to you know talk about the advantages of numpy and so on. No one and no from myself. I'll wait. I have time. No problems. I'll wait. But I want you guys to practice. And I insist. Guys, give that a try. If you're feeling not comfortable with the code, I will you know show up whatever maximum possible maybe you might require this line as well. Paste up the results of a time before it took up and time after. So Rosie it took up 0.013 01 and 0.00 potentially Rosie is having either a very good machine or she's using Google Collab. She or he is using Google Collab. I'm not pretty much sure. I cannot capture with the first name. Is it a he or a she? Sorry for that. Now that can be present as a rosie and can also be pronounced as rosé. So slight confusion. Not a problem. I am not gender biased. It's anything. I'm okay with that. But good potentially what I can inference from that either you have a very good machine or you're using Google collab servers if I'm not wrong. Can you confirm again data start off with length 10. Prii has tried to use no service slightly utilize the distance things but yeah you can still notice the difference in the time isn't it right you have used a small chunk of data that is why the time taken has not too much of difference but there is a difference right so essentially what happens is because of the numpy functionality Right? I can use or I can perform task in a faster manner as compared to the traditional approach. And that is why people prefer using numpy and not using the functions uh not using the default or core functions or core python part. Core python is good. We need to use certain things from core python. Great. But yet you would require libraries to do work with this. Now does anybody require help installing numpy onto the machine? Especially to the people who are using VS code and not able to extract uh numpy. Anybody not able to execute any of the piece of codes that I have discussed so far? So suggest some system for data science AI typically that's difficult to gather for me what kind of system that you're looking for thank you very much Narendra on that that part install empire minimum requirement see there is uh even if you have a very basic machine right you have a basic machine that allows you to you know browse internet over any period of time not a problem what you want I want you guys to have a good browser right and you can still use Google collab I'm not saying that you need to have a locally installed software not necessarily you can use the cloud softwares the point is we need to be a little bit more you know conscious with them because they lose out on the data I need to keep a local copy of those data so a good a simple average machine where any simple average machine like let's say you can have a 4 GB of RAM with any simple processor I'm okay with that but make sure you have a good internet speed right if I talk about my machine right now I'm working on currently on 200 megabytes 200 Mbps connection right now that's the maximum available in my area if I would have some more options I would have gone with So good internet speed works well provided you have an average machine which is good enough to play you know videos watch out anything I'm okay with that. So nothing specific from my end a minimum requirement but yes something that you can have that allows you to do a proper you know proper without lag functionality I'm okay with that 2 GB of RAM is also fine with me I'm okay with that provided your browser works well great now comes the first aspect that is the array now before I move on to the arrays can Someone tell me the list. Anybody heard about the word list before list in Python? No. Great. Okay. Heard about ling list in other languages? Now data structure an ordered mutable collection that stores multiple values in one variable. Thank you sesh. So essentially understand the idea Python offers a data structure. Data structure is an arrangement. Right? It offers a data structure which is called as a list. Now list can contain any type of value. Right? I can have a integer value. I can have a floating point value. I can have any sort of I can have a string. I can have any boolean type of data. The point is when the list works, what does it do? It tries to store your data into the memory. Now it does not store that into a continuous manner, right? It will try to store integer value somewhere, a floating value somewhere else or another value somewhere else, somewhere else. Now whatever the combination it can be you know continuous it may be continuously together it may not be together it it varies machine to machine right wherever the memory is available a list will store all data points accordingly now however the arrays allows me to store the same information in a in a contiguous manner that means that essentially if I have four or five or 10 number of values uh a array will allow me to store the data in a continuous manner like this. I have one 2.4 let's say letter A and true. But there is a catch. The catch is with list every individual element right of that data structure will keep its data type. That means if something is integer that will remain integer. If something is floating that will remain floating. If something is string it will remain string. And something is a boolean value will remain boolean. However with the arrays the data type changes. That means it is going to be throughout the array every element is of same data type. Every element of the array is of same data type. Are we good? Thank you. So what is an array? So an array I've given you the definition earlier to an array is a contiguous arranged data arranged data in memory right a continuously arranged data That's a simple array. Numpy offers numpy has array by the name or that's on numpy addresses. Let me just put that addresses arrays. as the name and the name is ND array which refers to n dimensional array which refers to n dimensional array what is the idea behind it so the arrays could be in multiple number of dimensions when When I say dimensions, what does it mean? It means that suppose I have a single line of data. It can be horizontal, it can be vertical, does not make a difference. I'll call that as 1D on one dimensional. If I have a table like this, I have rows and I have columns, right? Like this, I'll say that as two-dimensional. If I have a threedimensional data, right? I have rows and columns like this. behind that I have another layer like just like you can say that a Rubik's cube right you can think of a Rubik's cube or any threedimensional object right so if the data is arranged in that format I'll say that as a threedimensional data or hypothetically you can think of four dimension five dimension fifth dimension these are not pictorially representable but imaginary right but yeah we can represent them in terms of machine great so Python offers two ways is there are two ways. There are two ways by which the arrays can be created in Python. Right? What is the number one way? using the D using the array function. There's a function called as an array and then there is using some using the built-in array generative function. There are a lot of many of those. So I will not talk about them in uh in general. I'll try to talk about them one by one. Good. So let's try to create arrays. Let's create arrays. Create ND array rather using array function. Great. Now the first one that we have in the picture is let's create a zerodimensional array. Now what is zero dimension means input stream. What do you mean by zero dimension? When you see the word zero dimension, anything strikes back. Yes. Does not exist. No, it does exist. Priy a zero dimension does exist. Empty. No ro uh rosé roishand empty array I will yeah you can say that as zero dimensional but uh that indeed does not mean the same thing understand what is do we know what is the difference between a scalar data and a vector data I will though talk about scalar and vector in detail later on that's a separate thing but what is a scalar data and what is a vector data A scalar is without direction. Great. Thank you very much. Direction means a dimension.