Darth-Glennium/sierra-leone-spatial-demographics

An end-to-end reproducible R pipeline ({sf}, {tidyverse}) to audit, clean, and map subnational population dynamics and spatial densities for Sierra Leone (2020–2025). # Sierra Leone Subnational Population Validation & Spatial Analysis (Admin 2) An end-to-end reproducible data engineering and geospatial analysis pipeline implemented in R to clean, validate, and analyze subnational demographic trends across the 14 official administrative level 2 (Admin 2) units of Sierra Leone from 2020 to 2025. The primary objective of this project is to bridge routine tabular statistical reporting with exact geographic boundaries, ensuring a structurally sound, audit-ready dataset for downstream spatial modeling and resource allocation. ## Technical Approach & Workflow Architecture The analysis is contained entirely within a portable, production-grade R Notebook (`Population analysis - Glenn Kipanga.ipynb`) which moves sequentially through a defensive data engineering lifecycle: 1. **Environment & Path Isolation:** The environment dynamically provisions local user package libraries and leverages a multi-directory look-ahead check. This structure ensures full portability; the pipeline runs seamlessly whether executed from a centralized repository root or an isolated localized workspace. 2. **Data Ingestion & Attribute Auditing:** Schema geometries and tabular monitoring records are systematically scanned. Key reporting inconsistencies and structural OCR noise (such as alphabetical character `'O'` injections substituting numeric zeros, e.g., `"6O1994"`) are programmatically captured using selective regular expressions rather than blunt type coercions. 3. **Missing Value Imputation:** A total of 30 missing or zeroed temporal rows across the longitudinal sequence (2021–2024) are estimated. The pipeline drops identified data-entry anomalies and fits localized linear regression models (`lm(population ~ year)`) grouped per district. Predictions are strictly lower-bounded using a `pmax(..., 0)` truncation to enforce demographic validity. 4. **Administrative Reconciliation:** String-cleaning primitives standardize cases, drop trailing white spaces, …