This repository contains experiments and projects exploring advanced Python concepts and libraries. It is a work-in-progress for learning and practicing new Python skills.
- Email Handling: Sending and reading emails using
smtplib,imapclient, andpyzmail. - Security: Handling passwords securely with
getpass. - GUI Automation:
- Controlling the keyboard and mouse using
pyautogui. - Automate repetitive tasks like filling out forms, pressing buttons, and navigating applications.
- Use screen color checks and screenshots to help the program stay on track.
- Best practice: make programs fail quickly on invalid instructions to prevent unintended actions.
- Simulate human-like actions and watch the mouse move while text appears automatically.
- Controlling the keyboard and mouse using
- Data Analysis & Visualization:
- Analyze COVID-19 datasets using
pandas. - Visualize global confirmed, recovered, and death cases with interactive Plotly charts.
- Enhanced visualizations with styled pie/donut charts, annotations, and color-coded segments.
- Analyze COVID-19 datasets using
- Natural Language Processing & Deep Learning:
- Build deep learning models using
TensorFlowandKeras. - Text preprocessing with tokenization, padding, and GloVe embeddings.
- Convolutional layers (Conv1D) to detect local patterns in text.
- LSTM layers to capture long-term dependencies in sequences.
- Binary classification models for tasks like fake news detection.
- Build deep learning models using
- Reads a news dataset containing titles, text, and labels (FAKE/REAL).
- Preprocesses data by combining title and text, tokenizing, and padding sequences.
- Encodes labels into numerical format for model training.
- Uses pre-trained GloVe embeddings to represent words in real-valued vector space.
- Model architecture:
- Embedding Layer: Maps words to dense vectors using GloVe embeddings.
- Conv1D + MaxPooling: Detects local textual patterns and highlights important features.
- LSTM Layer: Captures context and long-term dependencies in the text.
- Dense + Sigmoid: Outputs final classification probability (Fake or Real).
- Trains on a subset of the dataset and validates on a test set.
- Makes predictions on new articles by combining title and text, tokenizing, padding, and feeding to the trained model.
- Saves the trained model for reuse.
- Provides tokenized sequences and prediction probability for transparency.
- COVID-19 Global Analysis with Plotly:
- Reads time-series CSV datasets for confirmed, recovered, and death cases.
- Computes total active, recovered, and dead cases.
- Generates interactive donut charts with custom colors, annotations, and hover information.
- Produces shareable HTML chart files for visualization.
- Programming Language: Python 3.x
- Data Analysis & Visualization:
pandas,numpy,plotly,matplotlib,seaborn - Machine Learning & Deep Learning:
tensorflow,keras,scikit-learn - Text Processing & NLP:
keras.preprocessing.text.Tokenizer,pad_sequences, GloVe embeddings - Automation & GUI Interaction:
pyautogui - Email Handling & Security:
smtplib,imapclient,pyzmail,getpass - File Handling & I/O: CSV, HTML export for interactive visualizations
The goal of this repository is to gradually build experience with advanced Python functionality, including:
- Natural language processing and deep learning model development
- Automation of repetitive tasks on your computer
- Email interaction and handling
- Secure handling of sensitive information
- Data analysis and interactive visualization
New projects and experiments will be added over time as skills progress.