Before data can be modeled, it must be refined. This repository showcases automated techniques for cleaning and restructuring raw datasets, ensuring they are accurate, consistent, and ready for machine learning or visualization.
- Focus: Automating the "cleanup" of messy spreadsheet data.
- Implementation: Using Python (Pandas/Openpyxl) to handle merged cells, inconsistent date formats, and null values.
- Key Operations: Pivot table automation, column splitting, and multi-sheet consolidation.
- Focus: Domain-specific data wrangling for the entertainment industry.
- Implementation: Processing a raw movie theatre dataset involving ticket sales, showtimes, and snacks inventory.
- Key Operations: * Handling Categorical Data: Encoding genres and ratings.
- Time-Series Prep: Converting timestamps into features like "Peak Hours" or "Weekend vs. Weekday."
- Outlier Detection: Identifying and correcting anomalies in pricing or attendance records.
- Language: Python
- Libraries: Pandas, NumPy, Openpyxl
- Tools: Jupyter Notebooks / CSV & Excel Engines
- Data Cleaning: Dropping duplicates and filling missing values using statistical imputation.
- Feature Engineering: Creating new variables from existing data to enhance model performance.
- Normalization: Scaling numerical data for consistency across different scales.