An automated data preprocessing and cleaning pipeline built using Python and Pandas to transform unstructured, messy global COVID-19 records into clean, analysis-ready datasets.
- Data Auditing: Inspected shapes, data types, and structural null gaps using Pandas profiling tools and Seaborn.
- Granular Extraction: Separated localized country records by filtering out macro-regional summary totals (like 'Europe', 'Asia', 'All') and cruise ship entries.
- Null Imputation Strategy: Handled missing tracking attributes reliably without biasing numerical means, implementing explicit Pandas nullable integer formatting (
Int64). - Text Normalization: Standardized mixed-cased typographic naming entries through clean structural strings optimization patterns (
.str.strip(),.str.title()).
- Python 3.x
- Pandas: Advanced indexing (
.loc/.iloc), string methods, and type-casting.