Box-Jenkins ARIMA Time Series Analysis of Malaria Deaths — General Hospital Toro, Bauchi State, Nigeria (2000–2024)
# Time Series Analysis of Malaria Deaths
### General Hospital Toro, Bauchi State, Nigeria (2000–2024)
---
## Overview
This repository contains a comprehensive **Box-Jenkins ARIMA** time series analysis of annual malaria deaths recorded at **General Hospital Toro, Bauchi State, Nigeria** from 2000 to 2024.
All statistical output is presented in **EViews format**, making it directly comparable to results generated by the EViews econometric software.
---
## Research Summary
| Item | Detail |
|------|--------|
| **Variable** | Annual Malaria Deaths |
| **Source** | General Hospital Toro, Bauchi State |
| **Sample** | 2000–2024 (25 observations) |
| **Method** | Box-Jenkins ARIMA (ADF & PP Unit Root Tests) |
| **Best Model** | ARIMA(0,1,2) — Minimum AIC & BIC |
| **Forecast Horizon** | 2025–2029 (5 years) |
| **Significance Level** | 5% (α = 0.05) |
---
## Key Findings
- **Stationarity**: Series is I(1) — confirmed by both ADF and Phillips-Perron tests across all specifications
- **Model Selected**: ARIMA(0,1,2) with AIC = 176.2495 and BIC = 179.7836
- **Diagnostics**: All Ljung-Box Q-statistics (lags 1–12) indicate no residual autocorrelation — model is well-specified
- **Forecast**: Projected deaths of approximately 52–54 per year from 2025–2029
---
## Repository Structure
```
malaria-time-series-analysis/
│
├── data/
│ └── Time_series_research_dataset.xlsx # Raw data from General Hospital Toro
│
├── output/
│ └── Malaria_ARIMA_EViews_Analysis.docx # Full EViews-style report (Word)
│ └── Malaria_ARIMA_EViews_Analysis.xlsx # Full EViews-style report (Excel)
│
├── code/
│ └── arima_analysis.py # Python analysis script
│
└── README.md
```
---
## Analysis Contents (in Report)
1. **Data & Descriptive Statistics** — Mean, Std. Dev., Skewness, Kurtosis, Jarque-Bera
2. **Unit Root Tests** — ADF & PP tests at level and 1st difference (all 3 specifications each)
3. **Correlogram** — ACF & PACF at level and 1st difference
4. **Model S …