So I’m pivoting → StockSimPy will be beginner-friendly:
📚 Students & newcomers
⏱️ Backtest in 5 lines of Python
💰 See profits/losses instantly

Full write-up here: medium.com/stackademic/...

Other libraries already do that — but they’re too confusing for beginners.

Sometimes you just want to test:
SMA50 > SMA100

You shouldn’t need 200 lines of code for that.

Curious how others handle project documentation. Do you automate it? Skip it? Obsess over it? Let’s chat.

#Python #OpenSource #Sphinx #DevDocs #buildinpublic #StockSimPy

I broke it all down in this post:
📚 “Setting Up Proper Documentation with Sphinx — Building StockSimPy #1”
👉 medium.com/python-in-pl...

The best part? Once set up, adding new modules and functions to your docs takes just a few lines of config.

It’s a game-changer for long-term maintainability.

Why spend time on docs?

They make your code usable by others (and your future self)

They boost credibility—especially for open-source work

And with Sphinx, maintaining them doesn’t feel like a second job

In my latest blog post, I cover:
✅ How to set up Sphinx in minutes
✅ Autogenerating docs from your code with autodoc
✅ Using Google-style docstrings and napoleon
✅ Making your docs future-proof and contributor-friendly

Excited to share more projects like this soon.
Follow if you're into #AI, fast builds, or ML deployment 🚀

Full article with ROC curves, ML decisions, and takeaways:
👉 blog.stackademic.com/diaguide-how...

Streamlit Cloud hosted the app.
All I needed was a GitHub repo and a requirements.txt.
Push and deploy!

The app was built in Streamlit.
→ No frontend coding
→ Fast and responsive
→ Easy deployment

I tested 3 ML models:
– Logistic Regression
– Random Forest
– Gradient Boosting

Logistic Regression had AUC ~0.81 and was fastest, so I picked it.

I cleaned a public dataset and chose to remove income and education, which only improved accuracy slightly.
The focus was on trust and accessibility, not just numbers.

The goal:
– Predict diabetes risk
– Use ethical, transparent modeling
– Deploy it on the web

I called the project DiaGuide.