https://github.com/nicolaspetrov/python-projects

This folder contains all my Python projects.
https://github.com/nicolaspetrov/python-projects

machine-learning ml python telegrambots

Last synced: 10 months ago
JSON representation

This folder contains all my Python projects.

• Host: GitHub
• URL: https://github.com/nicolaspetrov/python-projects
• Owner: NicolasPetrov
• Created: 2025-08-20T13:47:42.000Z (10 months ago)
• Default Branch: main
• Last Pushed: 2025-08-20T14:36:25.000Z (10 months ago)
• Last Synced: 2025-08-20T16:27:01.430Z (10 months ago)
• Topics: machine-learning, ml, python, telegrambots
• Homepage:
• Size: 4.88 KB
• Stars: 1
• Watchers: 0
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

          
# Python-Projects

Welcome! This folder contains all my Python projects. Below are brief descriptions and links to the corresponding folders.

## 1) Task Manager - Personal Task Manager



A **personal task manager** with a Flask web UI and **EN/RU localization**. Demonstrates a production-style CRUD app: modular architecture, internationalization, error handling, and an easy deploy setup. Suitable as a starter for learning projects, pet tools, or small internal apps.

**Key Features**

- Tasks: create, edit, complete, delete

- Priorities: **High / Medium / Low**

- Statuses: **Pending / In Progress / Done**

- Due dates with overdue highlighting

- Modern, responsive UI (Bootstrap 5)

- Internationalization (EN/RU, EN by default)

- Storage: SQLite (single DB file) with a simple migration path

**Technical Details**: Python 3.7+, **Flask**, **SQLAlchemy**, **Flask-Babel**, Bootstrap 5, Jinja2  

**Structure (main)**: `app.py`, `run.py` (port 5001), `requirements.txt`, `templates/`, `translations/`

Learn more →

---

## 2) Housing Price Predictor



A **production-ready ML application** for real estate price prediction using **XGBoost**, interpretability (SHAP/LIME), and a **Streamlit web interface**. The project demonstrates the full cycle: data generation/preparation, training with hyperparameter tuning (Optuna), artifact persistence, dashboards, error handling, and a modular architecture. Supports **EN/RU** interface switching.

**Key Features**

- XGBoost + Optuna (10k+ records, automated training)

- Streamlit UI: real-time predictions, data analysis, visualizations

- Interpretability: **SHAP** (feature importance), **LIME** (local behavior)

- Advanced preprocessing: features for location/infrastructure/environment, outlier handling (IQR)

- Metrics: R², RMSE, MAE, MAPE + plots

- Logging, robustness to missing values, structured persistence of models/configs

**Technologies**: Python 3.8+, `xgboost`, `streamlit`, `shap`, `optuna`, `plotly`

**Structure (main)**

- `app.py` — web interface (EN/RU)

- `train_model.py` — automated training & optimization

- `config/config.py` — data/model parameters

- `src/model.py` — XGBoost wrapper (fit/predict/eval)

- `src/data_processing.py` — preprocessing & feature engineering

- `src/visualization.py` — plots & dashboards

- `src/explainer.py` — SHAP/LIME

Learn more →

---

## 3) GifAnimalBot



A Telegram bot built with **aiogram 3.x** that sends random animal GIFs from the **GIPHY API** (cats, dogs, capybaras, parrots, pandas, otters). It includes **usage statistics**, a **size filter (<5MB)** for Telegram compatibility, inline buttons, and resilient error/retry handling. A solid example of an asynchronous bot with a modular architecture.

**Key Features**

- Choose an animal → random GIF

- `/stats` — counter of received GIFs and favorite animals

- Size filtering (<5MB) based on GIPHY size metadata

- Inline buttons: “More”, “Another Animal”

- In-memory GIF caching; de-duplication via JSON persistence

- Network retries and logging

**Technical Details**: Python 3.13, `aiogram` (3.x), `aiohttp`

**Structure (main)**

- `config.py` — tokens (Telegram/GIPHY)

- `gif_manager.py` — fetching/caching/size filtering/persistence

- `keyboards.py` — inline keyboards

- `handlers.py` — commands & callbacks (`/start`, `/stats`)

- `bot.py` — entry point, polling with retries

- `used_gifs.json` — prevents duplicates

Learn more →