I built a synthetic HIV dataset (500 participants, 81 variables) modeling adolescent and young women in rural Nigeria. I created it using R using AKAIS 2017 and NDHS 2018 data, it supports analysis, modeling, and public health research where real data is limited.
# HIV-in-Akwa-Ibom-Analysis-with-R
I built a synthetic HIV dataset (500 participants, 81 variables) modeling adolescent and young women in rural Nigeria. I created it using R using AKAIS 2017 and NDHS 2018 data, it supports analysis, modeling, and public health research where real data is limited.
Overview
This project generates a synthetic (computer-simulated) dataset for HIV research in rural Akwa Ibom State, Nigeria. The data is NOT from real people, but the patterns and probabilities are based on actual research findings from:
AKAIS 2017 - Akwa Ibom AIDS Indicator Survey (1,818 adolescents)
NDHS 2018 - Nigeria Demographic and Health Survey (1,068 sexually active youth)
Installation
bash# Clone this repository
git clone
github.com
# Navigate to the directory
cd akwa-ibom-hiv-dataset
Generate Your Dataset (3 Steps)
Open R or RStudio
Run the script:
r source("generate_hiv_dataset.R")
Export Your Data
r# Save as CSV (for Excel, SPSS, Stata)
write.csv(hiv_data, 'my_hiv_dataset.csv', row.names = FALSE)
# Save as RDS (for R)
saveRDS(hiv_data, 'my_hiv_dataset.rds')
Dataset Details
The Basics
FeatureDetailsSample Size500 participantsLocation2 rural LGAs (Ika, Ibiono Ibom)Age Range15-24 yearsGenderFemaleVariables81 total (75 primary + 6 derived)FormatR data frame, exportable to CSV/SPSS/Stata
What's Inside
The dataset covers all major areas of HIV research:
1. Demographics (8 variables)
Age, education, marital status, occupation, income, religion, household size
2. HIV Knowledge (10 variables)
Transmission knowledge
Prevention knowledge
Misconceptions (mosquitoes, food sharing, witchcraft)
Finding: Only ~2% have comprehensive HIV knowledge (matches literature: 9.4%)
3. HIV Testing (5 variables)
Testing history and frequency
Status awareness
Willingness to test
Finding: ~29% ever tested (low coverage)
5. Sexual Behavior & Risk (9 variables)
Sexual activity
Age at sexual debut
Number of partners
Condo …