Emane-11/maji-ndogo-agricultural-analytics

# Maji Ndogo Agricultural Analytics, Data Pipelines & Hypothesis Validation ## Project Context This project focuses on identifying the environmental and management drivers behind crop yields across farms in the fictional region of Maji Ndogo. The dataset captures detailed parameters on field sizes, crop classifications, field elevation, recorded rainfall and temperature, and pollution levels. However, cross-referencing these manual field survey logs with automated weather station networks revealed a significant data quality hurdle: a substantial portion of the climate data recorded by field auditors deviated from physical sensor baselines. To build a reliable foundation for agricultural planning and predictive modeling, this project: 1. **Discovers and Maps Variables:** Profiles the relationships between environmental factors (such as rainfall and elevation) and agricultural output (`Annual_yield`). 2. **Automates the Engineering Pipeline:** Refactors step-by-step cleaning logic into robust, object-oriented code modules to handle incoming data updates. 3. **Validates Findings Experimentally:** Switches from arbitrary tolerance margins to strict **Two-Sample Independent T-Tests (Welch's T-Test)** to mathematically verify whether survey deviations represent acceptable sampling variance or structural data errors. ### ⚠️ Acknowledgements & Context * **Educational Context:** This portfolio project was developed as a comprehensive extension of the **ALX Data Science / Data Analytics Integrated Project series**. * **Data Provenance:** The underlying datasets are sourced from **ExploreAI**. * **Data Authenticity:** All records, schemas, geographical locations (Maji Ndogo), and survey parameters are entirely **fictional** and used solely for data engineering, statistical validation, and agricultural analysis modeling. ## 📂 Repository Structure The architecture of this repository decouples reusable data engineering modules from local analytics, visualization, and v …