README.md

AHRI R library

The ahri R library provides functions for estimating the HIV incidence rate with Africa Health Research Institute (AHRI) data. These functions can read in the AHRI datasets, write them to .Rda format, standardize and subset the data, impute the HIV infection events, and calculate unadjused or age-adjusted HIV incidence rates using a single or multiple imputation approach.

The wiki help pages serve as a short introduction to the ahri library. These can be found in the links below. The help files are organised as follows:

• Getting started, which describes how to install the ahri library, which AHRI datasets to request and where to put them. It also shows how to set the paths to these datasets. https://github.com/vando026/ahri/wiki/1-Getting-started

• Reading and writing the datasets, which describes the functions for performing these operations. https://github.com/vando026/ahri/wiki/2-Read-functions

• Set functions, which describes a range of functions for processing the data, subsetting the data, and other data transformations. https://github.com/vando026/ahri/wiki/3-Set-functions

• Utility functions commonly used with HIV incidnece analyses. https://github.com/vando026/ahri/wiki/4-Utility-functions

• Functions to make the HIV incidence datasets, impute the HIV dates, perform mid-point, end-point, or single random-point imputation, and calculate the HIV incidence rates by sex or year using multiple imputation. https://github.com/vando026/ahri/wiki/5A-HIV-functions and https://github.com/vando026/ahri/wiki/5B-HIV-functions

• G-imputation functions to impute the HIV times conditional on auxiliary data. https://github.com/vando026/ahri/wiki/6-G-Imputation

There are other sources of help:

• The ahri package has help files and documentation. Type ?ahri to get to the help pages. For more information on a specific function, for example setFiles, type ?setFiles.

• Please consult the issues page on this Github site for more information and for answers to questions someone before you may have already asked.

• There is a Python version of this library that speads up the HIV incidence calculations.

Screen shots

Read and load the data, subset and create repeat-tester data.

Calculate the HIV incidence rate for women aged 15–24 years. Do 3 imputed datasets using the single-random point method.