
Exploring Data with Python
Chapter 1
Getting Started with Python and Jupyter Notebooks
Michael Thompson
Hey everyone, welcome back to the CSC108-Week-3 Podcast. I'm Michael Thompson, and today we're gonna dig into the essentials of data exploration with Python. So, if you're just starting out, or maybe you’ve opened up Jupyter Notebooks for the first time and thought, “Uh, what do I do now?”—you’re in the right place. I remember my first time using Jupyter Notebooks in a classroom setting. I’ll be honest, it was a bit intimidating. I mean, you open it up and there’s this blank cell staring at you, and you’re supposed to just start typing code? I kept wondering, am I supposed to run this whole thing at once, or just one cell? And then, of course, I accidentally deleted a cell and panicked for a second. But, you know, that’s part of the learning curve.
Michael Thompson
So, let’s break it down. Python is a great language for beginners—very readable, not a lot of weird syntax. And Jupyter Notebooks, well, they’re kind of like a digital lab notebook. You can write code, run it in little chunks, and see the results right away. That’s super helpful when you’re experimenting or just trying to figure out what works. The practical workflow is pretty straightforward: you write a bit of code in a cell, hit Shift+Enter, and boom, you see the output right below. You can organize your code into sections, add notes, and even include charts or tables. It’s all about making your work reproducible and easy to follow, especially when you come back to it later and think, “Wait, what was I doing here?” Trust me, that happens to all of us.
Chapter 2
Mastering Descriptive Statistics
Michael Thompson
Now, once you’re comfortable running code in Jupyter, the next big step is actually exploring your data. And that’s where descriptive statistics come in. You’ve probably heard terms like mean, median, and mode thrown around. Let’s break those down real quick. The mean is just the average—add up all your numbers, divide by how many you’ve got. Median is the middle value when you sort everything, and mode is the number that shows up the most. Simple, right? But these little numbers can tell you a lot about your dataset.
Michael Thompson
Let’s say you’ve got a dataset of exam scores. If you run a quick mean calculation in Python—something like sum(scores) divided by len(scores)—you get a sense of the overall performance. But maybe there’s a big gap between the highest and lowest scores. That’s where median helps, especially if you’ve got outliers. And if a bunch of students all scored the same, mode will point that out. In Jupyter, you can use libraries like pandas to do this in just a couple lines of code. For example, you can call .mean(), .median(), or .mode() on a pandas DataFrame column, and it spits out the answer. Super handy.
Michael Thompson
But why does this matter? Well, descriptive statistics help you spot patterns and outliers. Maybe you notice the mean is much higher than the median—that could mean you’ve got a few really high scores pulling the average up. Or maybe the mode is way lower than the mean, which might tell you most students struggled with a particular test. These insights are the first step in understanding your data before you dive into more complex analysis. And honestly, sometimes just looking at these basic stats can reveal things you’d never notice by just scrolling through rows and rows of numbers.
Chapter 3
Comparing Data Exploration Tools
Michael Thompson
Alright, so you’ve got the basics down in Python and Jupyter, but maybe you’ve heard about these low-code tools like Weka or Orange. How do they stack up? Well, let’s do a quick side-by-side. In Python, you’ve got a lot of flexibility. You can write custom code, automate tasks, and handle pretty much any dataset you throw at it. But, it does mean you need to know a bit of programming. With low-code tools, you get a more visual interface—drag and drop, click a few buttons, and you’re running analyses without writing much code, if any.
Michael Thompson
For example, if you just want to quickly visualize your data or run a simple classification, Orange makes it really easy. You can connect widgets, load your data, and see results almost instantly. But if you need to do something more custom—like, say, a specific data transformation or a unique visualization—Python’s your best bet. I had a student last semester who actually used both approaches on the same project. They started with Orange to get a feel for the data, ran some quick visualizations, and then switched to Python when they needed to dig deeper and automate some of the analysis. It was a great example of how these tools can complement each other. There’s no one-size-fits-all answer—it really depends on what you’re trying to do and how comfortable you are with coding.
Michael Thompson
So, whether you’re a Python purist or you like the convenience of low-code tools, the important thing is to keep exploring and experimenting. That’s how you really learn what works for you. Alright, that’s it for today’s episode. Next time, we’ll dive even deeper into data analysis techniques, so stay tuned and keep practicing. See you then!