Patrick Fromaget (RaspberryTips)’s Post

AI Engineer: Day 2/100 Extract text from a PDF document using Python. Remove stop words (useless words like "and," "with," or "or"). Convert the remaining text into tokens (tokenization). Library: PyPDF2 Benefit: Document parsing Comment below if you want the code!

Every tutorial, every YouTube video, every “Build your first AI app” article starts with pip install. That creates a false impression — that Python isn't just popular, but required to build with AI. It isn’t. Here is my article bursting the Myths around AI, Python and Java https://lnkd.in/g8-z9tZN

Small reminder from today’s Python learning I finally understood that in a function like: def add_vehicle(vehicles, new_vehicle_params): “vehicles” is just a parameter A temporary name for whatever list you pass into the function. So when you write: add_vehicle(existing_vehicles, new_vehicle) Python reads it as: vehicles = existing_vehicles Such a small thing, but understanding it made functions feel way less confusing. Sometimes growth in tech is just understanding one small concept at a time ✨ #Python #DataAnalytics #LearningInPublic #WomenInTech #TechJourney #alx #BeginnerProgrammer

Built Linear Regression from scratch using Python (no libraries) Wanted to understand what’s happening under the hood before moving to sklearn. So I implemented a simple model to predict marks based on hours studied using Gradient Descent. 🔹 What I did: Implemented the prediction function (y = wx + b) Calculated Mean Squared Error (MSE) manually Computed gradients and updated parameters over 1000 epochs 🔹 What I learned: How gradient descent updates weights step by step Why learning rate plays a critical role How loss decreases as the model learns 🔹 Result: The model successfully learned the relationship. Example: If a student studies 9 hours → predicted marks ≈ 89.3 🔗 Code: https://lnkd.in/gPHCenhB Next step: implementing this using NumPy and then sklearn. #MachineLearning #Python #LearningInPublic

🚀 Day 6/30 – Python Challenge Exploring loops in Python today! 🐍 🔹 Key Concepts: * for loop using range() * while loop execution * Iteration and repetition in programs 💻 Mini Task: Printed numbers from 1 to 5 using both for loop and while loop to understand their working. 🎯 Learning Outcome: Learned how loops help automate repetitive tasks and make code more efficient. Consistency + practice = improvement 📈 #Python #CodingChallenge #LearningJourney #AI #StudentDeveloper #Day6

🐍 Python Trick — Did you get it right? b = a doesn't copy the list. It points to the SAME object in memory. So when b changes... a changes too. 🤯 This one gotcha has caused more bugs than most people admit. 💡 Always use b = a.copy() or b = a[:] when you need a true copy. Drop a ✅ if you got it right or a ❌ if it surprised you! Follow for more Python tricks, AI/LLM tips & SQE insights every week. 🔔 #Python #PythonTricks #SoftwareEngineering #SQE #Coding #100DaysOfCode #AIEngineering #TechLinkedIn #PythonDeveloper

🚀 Built a small AI tool to chat with PDFs I made a project where you can upload a PDF and ask questions. It gives answers only from the document. If the answer is not in the PDF, it shows: “Not found in document”. 🔗GitHub : https://lnkd.in/gbs2ugHp 🛠 Used: Python, LangChain, FAISS, Streamlit Infotact Solutions #AI #Python #RAG

When I first started using Pandas, I wrote code the same way I wrote normal Python. Lots of loops. Lots of step-by-step logic. And it worked… at first. But then datasets got bigger. And things slowed down quickly. That’s when I learned something important: 👉 Pandas works best when you think in vectorized operations. Instead of: looping through rows You start thinking in columns. Example mindset shift: Instead of processing each row individually, you transform entire columns at once. This small change made my code: ✔ faster ✔ simpler ✔ easier to read Still learning, but it's one of those small mental shifts that really changes how you work with data. #DataEngineering #Python #Pandas

Python Clarity Series – Episode 25 Topic: Mutable vs Immutable Function Behavior 📌 Why did my list change after function call? def modify(lst): lst.append(100) a = [1, 2] modify(a) print(a) Output: [1, 2, 100] 👉 Lists are mutable → changes reflect outside Now: def modify(x): x = x + 10 a = 5 modify(a) print(a) Output: 5 👉 Integers are immutable → no change outside 💡 Rule: Mutable → changes persist Immutable → changes don’t This confusion causes logic errors. #PythonBasics #FunctionConcepts #StudentClarity #python #clarity

While learning LangGraph, one small Python concept suddenly became much more important to me: TypedDict. At first, I thought it was just “type annotations for dictionaries.” Useful, sure—but nothing special. Then I started thinking about state. When multiple nodes in a workflow keep reading and updating shared data, an unstructured dict becomes chaos very quickly. - Missing keys. - Unexpected values. - Confusing debugging. TypedDict solves that by forcing structure into state. That was my takeaway: - Sometimes tools that look “optional” become essential once systems start growing. #Python #BackendDevelopment #LangGraph #AIEngineering #BuildInPublic