Data Visualization with Matplotlib: From Basics to Advanced Techniques

1. Intro: Welcome to the MD Blot Leap course. In this course, we will learn how to work with Matlot Leap, the most popular and widely used library for data visualization and Python. It's highly demanded in the fields of data science, machine learning, and analytics because it makes creating powerful visualizations quick and easy. We will also cover the main functions and capabilities of Matplot Leap, including how to create different types of plots, customize them and make interactive visualizations. You will learn how to use MD Blot Leap effectively to present data in a visually compelling way. In our first project, we will create an animation using fake data that we will generate with a simple Python function and the num pi library. This will give you hands on experience with how animations work and how to dynamically update your plots. For the second project, we will use the Python library Y Finance to fetch real stock data for Google and Apple. We will then create an animated graph to visualize the stock price movements over a specific period of time, showing you how two real world data can be represented dynamically. By the end of this course, you will have a solid foundation in creating different types of plots and animations with Md plot leap and be ready to explore even more advanced features. Let's get started. 2. Matplotlib Setup & Basics: Line and Scatter Plot Tutorial: It's important to know MD Blot leap because it allows creating charts and diagrams for visualizing data, making it easier to understand and analyze information. By using MD Blot leap along with Bands, you can easily visualize data from a data frame. The library provides a wide range of settings and customization options for plots and diagrams. Before we begin, make sure you have Matlod lip installed. If not, use the Command Pep Install Matlod Lip. Or if you're using Anaconda, you can use this command. I already have Mud blood leap installed, so all I will do is launch Jupiter Notebook, and we will start working together. I'm opening the terminal, activating my virtual environment. Which already has Jubiter notebook installed, what a virtual environment and how to work with it, you can see in my bonus video. You don't have to use a virtual environment right now. You can comfortably work in the terminal. But for the future, knowing what it is will be very useful. Also, it's very convenient when you need to work with different projects and different versions of libraries. I will slightly increase the size of our terminal, and here I'm launching Jubiter Notebook, having previously navigated to the directory where our project will start automatically. You can use Jupiter Notebook in whichever way works best for you. We are right now in Jupiter in this directory where we will create our first file to explore Mat Blot Leap. Let's create our file. So let's go through the basic steps of using Matplot Leap. First, I import the library with the Alias PLT. Next, we need data to work with. Data based on which we will build our plot. Let's create something very basic like this. D. And now we can build our first data visualization, building the plot. And here it is, our first simple line plot. In the past, we used the magic command MD plot leaping line in Jupiter Notebook to display plots directly in the notebook. However, in the latest version of Jupiter, this command is usually not needed anymore. It's enabled by default. So when you create a plot, it will automatically appear in the notebook without needing to add Mud blot leap line. So now users can simply call visualization functions and the plots will automatically display. For now, we imported the Mult plot leap library and assigned the alias BLT, as is common practice for convenience. Then we created some data, not really fake, but data based on which we will build our plot. Then using the plot function, we created our plot. In the end, we either had to add Show function to display our plot or write Md Blot leap in line about it. But as I mentioned, in our case, we are working with the latest version of Jupiter notebook, and under the hood, we have this already. We don't need to take extra steps. I will remove this line because we don't need it. We can also add a label here and display it like this. And here it is. We can see our label on the chart. Now, let's build a scatter plot. Scatter plot is a powerful visualization tool that helps us understand the relationship between two numerical variables by plotting data points on two dimensional graph. Each point on the plot corresponds to one pair of values with the X and Y axis representing different variables. To create a scatter plot, we first need to pass the data for X and Y axis. These are the values that will define the positions of the points on the plot. Then we can specify the color, let it be red. Next, we specify the marker, which is the symbol or designation used to mark each data point on the plot. The marker can be of different types and size, allowing you to choose how you want to display individual data points. Then we indicate the label, which will be displayed in our legend, add the legend and here it is. This is what we got a scatter plot. 3. Exploring Data Visualization: Bar Charts, Multidimensional Analysis & Scatter Plot Styling: Now let's also learn how to build a bar chart. And to do this, I will slightly manipulate the database on which we will build this. Let's restart everything and build our bar chart. I will use the bar function this time. Then we pause our parameters again. Let's make it green and provide a label and of course, let's display it. A bar chart is a visual representation used to compare different categories or groups. They are commonly used in data analysis to represent data such as sales or any other categorical information. Now let's talk about how to add a title to our plot using the title function and how we can label our plot xs using X label and Y label. Let's go through this. For this, we will use the title X label and Y label methods. And look, now we get a title and labeled excess. Let's also learn how to change the style and color of our plot markers. Line styles and colors can help make our plot more informative. Let's experiment. For this, let's go to our line plot and make some changes. We just added a marker, set a style for our line, made the color blue, changed our legend. And we got a much better representation on the same line plot. Now, let's go back to the scatter plot. There's another parameter, S, which controls the marker size. We can set different values, experiment and see how it works. There is also a parameter called Alpha. It sets transparency. As you can see, we can experiment with this too. Zero means full transparency, and one is full opacity. Try experimenting yourself. Now let's work with multidimensional data. Scatter plots are not only for simple cases where you see the relationship between two elements. Let's imagine we have a dataset that contains three variables, and we want to determine how two variables interact and influence the third. For this, we will use a scatter plot and work with multidimensional data. So I import Numbi then generate random data that I will need. This line sets the initial value for the random number generator in the Numbi library. In the case of SED equals 42, the number 42 is chosen at the initial SED value. This results in the random number generator reproducing the same sequence of random numbers each time the program is run. Don't focus on it right now. It's not necessary for you to learn immediately. Just try to repeat it so you have an idea of what we are doing. Then we generate our multidimensional data, X one, X two, and Y. We generated random numbers using the Napi library, which is extremely important. If you haven't learned it yet, there is a tutorial in my profile. Check it. And here it is. Our data is ready. Now let's create a scat up plot based on it. We pass the first and second parameters, then the label. Here I will be looking at the interaction between X one and Y. Let's make the color blue, and then I will compare X two with the third parameter of Y. Labeling it with a different color. And here's what we have at the moment. Let's add a title for completeness. Then label the axis with X label and Y label. In this example, we use two sets of markers to display the relationship between X one and Y and between X two and Y. This way, we can easily compare the influence of both variables on the variable Y. Scatter multidimensional plots in MD plot leap are often used to visualize relationships between multiple variables. They help in understanding patterns, correlations or clusters within data. By representing each data point in multiple dimensions, they allow for better insights into complex datasets, often used in machine learning, data analysis, and statistics. 4. Exploring 3D Plots & Histograms in Data Visualization: We have more than two variables, three D plots become an important tool for visualization. Let's consider, for example, creating a three D plot for three variables. For this, we need to import a module from the Matplotlap library that contains the axis three D class designed for creating three D plots. This module extends the basic functionality of MDPlot leap for displaying and analyzing three dimensional data. It adds classes and functions that simplify the creation of various types of three D graphs. To avoid repetition, I will copy this here. Refine it a bit and rename variables. Now I will quickly write the code, and then I will explain what we did here. Meanwhile, post the video, copy the code, and then we'll go through it together. So we imported this module, then generated random data. Next, we create a figure object, which is the container for all the elements of the plot. It is a fundamental part of any Mdlolap plot. After that, we create a three D subplot using the AdSAlot method. This creates a subplot with a three D projection based on the figure object we created earlier. The parameter projection three D indicates that we want to create a three D subplot. After calling this command, we get an axis three D object that allows us to draw three D plots. We can call methods like scatter and others on this object to visualize three D plots. Let's call scatter using the scatter method and add points to the plot for X, Y, and Z. These are data for the respective dimensions. The C parameter is for color, so the color of the points is red. Then we specify the marker, which in our case is circular next, I specify the label to add a legend and then run all of this. Oops, we got a typo. Let's fix it. And look at the nice plot we got here. We can observe the scattering of plots in three dimensional space using the provided X, Y, and that data. This is very convenient when you need to understand relationships in complex datasets. Now let's learn about histograms. Histograms are an important tool for visualizing data distribution and identifying patterns in numerical datasets. They divide the range of values into intervals and display the count of observations that fall into each interval. So let's consider creating a histogram using MD plot leap and some random data. We generated data using Napi and in this example, we generate 1,000 random numbers. Then we call the hist function and pass our generated data. The Bins parameter determines the number of intervals. The color parameter sets the color of the histogram, and the edge color determines the color of the edges. Let's also add a title. Sorry for another typo. Let's quickly change that restart, and here it is all working. Let's add X label and Y label for a more user friendly display on the plot. And here's what we got. Pause this video and practice. Try to repeat or change something like changing the bean values or creating your own color for this histogram, or, for example, generating fake data to build a new histogram. You can customize every plot we build to your liking, and the more practice you have, the better. 5. Pie Charts, Saving Plots in Various Formats & Animations with FuncAnimation: A Pie chart is an effective way to visualize the proportions or frequencies of different categories in a dataset. This type of plot is particularly useful for representing portions, frequencies or percentages. Let's see how we can create a Pie chart. And first, I create data based on which we will build the Bie chart. In this example, I use the Pi function to create Bie chart. And here I pass the sizes. Parameter specifies the proportions for each category labels, parameter defines the names of the categories and the colors. Parameter sets the color for each category. Then comes the parameter that sets the format for displaying percentages. The start angle parameter allows rotating the chart at a certain angle. Usually it's set in degrees and can be any number 0-360. Let's add a title. The start angle parameter that in Mud blot lip specifies the angle from which the pie chart start drawing. By default zero degrees is typically at the top of the chart and without rotation. Drawing starts from the positive X axis direction, increasing counter clockwise. We can experiment with this and try different values, see how it works. Also, M plot leap provides an easy way to save plots in various formats, such as PNG, GPG, and PDF. This is especially useful when you need to use plots in other programs or publish them online. We already have our pie chart. Let's save it as a PNG file. You can also change the extension by specifying a different file format like GPG or PDF. The DPI parameter specifies the dots per inch resolution, allowing you to control the image quality. Run this cell. The file will be saved in the current working directory as hist PNG. Saving the plot may take a little time, and here we have our saved Pie chart. Bie charts are often used in marketing to illustrate the market share of products or services compared to competitors. They can also be used to evaluate the time spent on different project stages. Now let's dive into something more interesting Mud Blood leap animation, which allows us to build plots with live data in real time. Animation are useful for visualizing changes in data over time, making them ideal for real time data monitoring, simulations, and interactive storytelling in data science. We will learn how to create animations using funk animation from Md Blood leaps animation module. This function allows us updatablot at set intervals, simulating the effect of live data. I will switch to visualize Studio code for this example, as I find it more convenient for myself. So let's get started. First, we import everything we need. Of course, we will need some data. In real life, this data comes from external sources, for example, stock market charts or any other real time data visualization. In our case, we will generate the data ourselves. Let's create a function that will constantly generate new data for the plot. If you know Python basics, it will be easier to understand. But if not, don't worry, copy this function. Post the video and rewrite this code. I will explain what we are doing right now a bit later. In real project, you will receive data from external sources, so you don't need to fully understand this code right away. I hope that makes sense. So pause the video and rewrite this code. Right now, if I'm running the script, we will see a window with an animated plot. It might take a little time, but here it is. We can now see our data in real time. Notice this warning message. In some cases, the cache can grow unbounded, leading to potential memory issues, especially for long running animations or animations with a large amount of data per frame. To mitigate this issue, Dlod Lib disables frame data caching by default and displace this warning message. If you don't want to see this warning, you can follow the instructions and set cache frame data equals to false, or you can simply ignore it. Now let's break down the code step by step. First, we imported MD plot leap. Then we imported the funk animation class. We also imported Napi to generate random numbers. We need an update function that will be called to update the plot for each frame. In this case, the new data is generated randomly. But as I mentioned earlier, you can modify this part of the code to use real data from an external source. So what does the update function do? This function generates new data to be displayed in real time plot. We initialize a new data object that generates random numbers using Num Pi. Then we use a list in Python and the append method to add these generated numbers to the list, ensuring that the list is constantly updated. This continuously updating list forms the basis of our real time plot. Next, we have a line object, which represents the line being drawn in real time. We use the set data method on this object, passing two arguments. The first argument defines the X axis values, and the second argument defines the Y axis values. Here we use the length of the list as X values and the list itself for Y values. Then we create a new figure using PLT S alots. The function returns two objects, the figure and the axis. Next, we update the axis limits dynamically because the minimum maximum values of the X and Y axis change over time. To handle this, we use the Lin function. This recomputes the data limits based on current values. Then we have autoscale view. This automatically scales the plot so that all new data remains visible. This ensures that the plot smoothly updates as new data arrives. Now let's configure the properties of the plot. Here I used the legend function. You already know that this is added legend to a plot. It helps identify different elements by displaying labels for plotted data. I set X axis limits 0-50. Y aces limits 0-1. Then I set a title for the plot, and I add labels for both xs. Finally, to create an animation, I use funk animation. First, I pass the figure object to which the animation will be applied. Then I pass the update function, which will update each animation frame. The frame's parameter determines the number of frames. Since we set it to none, the animation continues indefinitely. The interval parameter determines the delay between frames in milliseconds. For example, if interval equals to 1,000, a new frame will appear every 1,000 milliseconds or 1 second. This controls the animation speed, and at the end, I call PLT show to display the animation. And this is what we got. There are many animation effects available in Mad Blot leap. You can see it in documentation. So you can experiment with them. Here you can find some more code. You can rewrite it and run. Feel free to add your own modifications. 6. Animating Stock Price Charts with Matplotlib & yFinance: Apple vs Google Comparison: Now let's do some practice. We will retrieve stock data for Apple and Google, then plot their stock price graphs over the past years and create comparative chart to determine which stock is currently more profitable for us. For this, I will use the Yahoo Finance and Mod Plot Lip libraries. Yahoo Finance is a library that allows you to obtain historical stock price data, dividends, balance sheets, and other financial indicators for publicly traded companies. First, we install the library using the package manager PIP in the previous example with funk animation, we simulated our data using the update function. Now I will fetch the real data from Yahoo Finance. After installing the library, we import it and retrieve Apple's stock data. I specify the ticker, which is a unique alphanumeric code, identifying a public company of the stock exchange. You can check available tickers here and choose the one you are interested in. In our case, we selected Apple. Next, I create an object and define the start and end dates for the period I want to analyze. Then using the history method, I retrieve historical stock price data for the selected period. Now let's bring the data we obtained. And here it is. It was so easy. Now we import Mtodlp and visualize the data. We create a figure. In matplot leap, a figure is a overall container that holds all the elements of a plot, including axis titles, labels, and the actual data visualization. It source as the canvas where multiple subplots or charts can be placed. The fig size parameter defines the size of the figure in inches. It controls how large or small the plot appears when displayed or saved. Adjusting fig size helps improve readability and layout, especially when working with multiple plots or detailed visualizations. Then use the plot function to draw a graph. As the first parameter, we have X access values, and here we have data index. The data object is a Pandas data frame that contains historical stock data. If you don't know what Pandas is, I highly recommend check my profile. You can find Pandas tutorial. This is extremely important library. So the index of this data frame typically contains dates time steps because Yahoo Finance provides time serious data. This means we are plotting stock prices over time with dates in X axis. The second parameter, data close, this represents the Y axis values. Close is a column in the data data frame, which contains closing prices of the stock for each date. The closing price is the last recorded price of the stock at the end of each trading day. I also added title and axis labels. And then display the plot using PLT Show function. The result is a graph showing changes in Apple's stock price over the selected period. Now let's add Google stock data and compare the prices of these two tech giant over the same period. We create a list of tickers, adding Google alongside Apple. Next, I create a list of objects, corresponding to each ticker using a list comprehensions. If you're not familiar with Ismprehension, welcome to my Python course. Just check out my profile here. There is an excellent course of Python that takes you from beginner to object oriented programming. I keep the start and end dates the same and create a dictionary that stores historical closing price data for each company. For this, I use a dictionary comprehensions. Quick tip. I highly recommend getting familiar with Python generators. You will often come across them, and they are very useful. Now let's adjust the layout a bit to make the code more readable. I remove print data. Next, I create a figure with a specified size, and I use for loop to plot the closing prices of each company's stock on the same graph. Each line represents a company's stock price trend. We set the title X aces label and Y Acess label and a legend using PLT legend function. And at the end, display the final comparative graph with PLT Show function. The result is the graph, comparing Apple and Google stock prices over the selected period, where each line represents the price movement of the company. Now let's make the chart animated. First, I import the animation module from MDPot Lip Then we create a few figure using FLT subplots and specify the axis labels. In the first case, when I used X label, ylabel, it was an attribute directly associated with the axis object. It's a way to access or modify the label of the X axis and Y axis. However, this is not recommended or most common approach. In this case, I use set X label and set Y label. The second case, these are the method provided by Mt Blot Leap to set the labels for the X axis and Y labels. It's a more robust and preferred way to interact with the plot. These methods is part of the official Mtplot leap API. So you just need to know that we have two variants. The first variant, these are attributes, and it might be a shortcut or older approach, but it's less flexible and less widely used. The same applies to the title. Next, we define the animate function, which will be called for each animation frame. This function clears the graph using X clear function and then plots the stock prices for each company by iterating through the data dictionary using the four loop. Each line in the animation corresponds to a different ticker, and we use plot to plot the data. The X axis represents dates and the Y axis represents closing prices. If you're familiar with Pandas, you already know that I alg selects rows or columns by index positions or range. Here, we dynamically select rows on each new animation frame, creating a gradual accumulation effect on the graph. The label parameter ensures that the correct company name appears in the legend. So as in the first case, we have data items, and from these data items, we got prices. And we have prices index as date, and we use it for X acess value. Then we use prices for the Y axis value. Now we create the animation using funk animation, similar to the previous example. Here I specify the number of frames the animation will have. It sets the total number of frames to the length of the data for the first ticker in the tickers list. Meaning the animation will update once for each data point in that specific dataset. This allows the animation to show how the stop price changes over time. We said the repeat equals to false parameter so that the animation doesn't restart after finishing. Finally, we display the animation. After running this code, we get an animated graph showing Apple and Google stock price changes over the selected years. Try experimenting with different stocks. Visit the NASDAQ website, choose other companies and create your own comparative chart. Which stock would you invest in? As you can see, Mat Bot Lip is a powerful library for data visualization and Python. It plays a key role in the Python ecosystem for scientific computing and visualization, providing flexible and efficient toolkit for creating a variety of plots, for simple charts, to complex visualizations for different analytical tasks. Mat Boot Lip remains one of the most widely used libraries in data science and finance. I highly recommend learn it. 7. Bonus: Efficient Work in Virtual Environment. Setting Up Your Workspace: Very often in reality, you will have to work with multiple versions of Python. This is because each project has its own technology stack and package versions. To avoid creating a mess on your work computer and dealing with conflicts between different versions, it's ideal to use a virtual environment. It's not urgently needed right now, but I suggest you to understand how it works. It will help you a lot. You can skip this part. It will not affect your learning of the Python basics. This will be more necessary when you start working on a project. And now let's get started. Guys, if you want to manage multiple Python versions on your machine, there is a tool PMF. It lets you easily switch between multiple versions of Python and change the global Python versions on your machine. Let's get started from the Macos and then I will show you how it works on Ubuntu. The first step you need to take before installing anything new, it's update. And just in case upgrade to upgrade all packages. The first command updates local repository metadata, the second command, Brew upgrade, upgrades all the installed packages on your system to the latest available versions. It's commonly used practice that users first run Brew update to get latest metadata and then run Brew upgrade to update their installed packages. This ensures that the system has the latest software installed. We go to the Github and follow the instructions. Then we use them Brew install for installing PMF. Just copy this command and execute it. Let's return to the documentation and see what we need to next. Scrolling down, and here we are I use Z SH or shell. It's a common line shell that serves as an alternative to the more commonly known Boss shell. So I copy all this code and post it in SHC file. So we have installed PNF, and now I want to talk with you about Virtual environment. Virtual Environment solves a real problem. You don't have to worry about packages you install on the main system location. PyNVirtoNv is a plugin for the PNF tool that allows user to create and manage virtual environments for Python projects. With this, you can isolate project dependencies more efficiently. So again, follow the instructions and install this plugin. After installation, we copy this command and add this to the SHRC file. In this case, we do it manually. Open the HRC file. It's a hidden file and typically located in the user's home directory. I use simple user friendly text editor nano. You can use VIM. Here we can see three rows of code that was executed when PAN was installed. And pause this command here. I write my comment for the better understanding in the future. Again, for nano text editor, I use the commands Control O and control exit. It allows me write and exit from the text editor. You can use your text editor and execute your commands. Then restart shell with this command. And we can use these tools. So let's check it. Here we can see a small documentation with commands for PNP and PM Vertls. The first command, we check PN version. It displayed the currently active Python version, along with information on how it was set. For now, I don't have any. If I want to now list all Python versions known to PM I use the Command Pimp versions. And for now, I didn't install any Python version with PMP. If I want to see the list of Python version available to be installed, I can use the Command PM install dash list. So let's try to install Python with PM. For this, we use the command PM install, and then I specify the version of Python. I will install another version of Python to demonstrate how you can work in isolated virtual environments with different Python versions. Now at the same time, you will learn how to install and remove Python using PNP. If I now check the versions, we will see several Python versions. The Asterix indicates that right now I'm in a system global environment, but I have two Python versions that I can use for creating new virtual environments for other projects. On this operating system, I have globally, I mean, Python version 3.10 eight. I said globally because for every project, we can have its own Python versions. And now with this command, I will create the first virtual environment. For test project, I use the command Py Virtual ENF. Then I choose the version of Python, and then I can call my virtual environment, whatever it's up to you. I will call it NF and version of Python. And now with the command PN virtual lens, I can see list all existing virtual environments. To activate my newly created virtual environment, I use the command PNF, activate, and then name my virtual environment V 3.90. I immediately can see that I'm in it. If I check the Python version here, it's 3.90, unlike the global version that we tested earlier. If you have several virtual environments and want to switch between them, you can execute the command PM, activate, and then name other virtual environment, even if you write now in another active virtual environment. Now if we install something here, it remains isolated from global environment. Any packages or dependencies installed inside my virtual environment will not affect the system wide Python installation or other virtual environments that we can create. So let's install something here. Let it be Jupiter. I go to the documentation and follow the instructions. Jupiter it's a tool for code execution. I choose Jupiter, for example, it can be any packages or libraries you want. Now with the command PIP freeze, I can see all packages that was installed in my virtual environment. PP is a package manager for Python. Now let's imagine that we don't need this virtual environment anymore. How we can delete it. First, if it's active, we should deactivate it with the command PM deactivate. Then we use the command Virtual delete and name our virtual environment that we want to delete. So when I check Pi on Virtual lens, we don't see our virtual environment anymore. It was deleted along with all packages and libraries that we installed there very useful thing. But it doesn't mean that Python version that we use in this virtual environment was also deleted. If we check Python versions, we still can see several. I added before another one, just to show you how we can uninstall Python versions with BMP, the command, uninstall and then version of Python and Viola, we uninstall Python version 3.9 0.8. With these tool, it's so simple manage different Python versions. Now let's install it on a Bunto. We do the same thing. Go to Github page and follow the instruction. Here, I chose automatic installer. And here I copy this command for installation. Before installation, I use the command pudo Ug update, this command, managing system packages, then psudoUg upgrade. So we update and upgrade all installed packages on our system. Now we can install PAMP Fast this command that we copy previously. Then return to the front page, or documentation, and copy these three lines of code. We will write it in Bahar C file. I is also hidden file. It's very similar that we did with MACOS previously. If you couldn't install Tmp and you got an error, make sure that you installed all dependencies for Python and have Gid on your PC. After all of this, restart shell with the command, and we can use this tool. We use here the same command we used previously on Macros. Let's install Python 3.90. Here we can see our installed Python. Now let's create virtual environment based on this Python version. We activate it the same way as previous with MacOS with the command and activate and then name a virtual environment. Probably you won't encounter this problem, but I have inconvenient behavior on my system. Right now, I don't see that I'm in virtual environment. If I check it, I can see that we have created it. So I had to add these few rows of code in my BachRCFle and everything works. On a Bundu, I also use nanotext Editor. These commands allow me to write and execute from the BachRCFle. Execute the command source BRC. It's for execution the BRC script in the current shell session. And while we fixed it. Right now we are in our virtual environment, and inside, we have the Python version that was used for creation. So, guys, all commands and all the next steps, the same as we did previously. I hope this knowledge will help you see you in the next lesson.