Learn R Programming

InSilicoVA (version 1.4.0)

Probabilistic Verbal Autopsy Coding with 'InSilicoVA' Algorithm

Description

Computes individual causes of death and population cause-specific mortality fractions using the 'InSilicoVA' algorithm from McCormick et al. (2016) . It uses data derived from verbal autopsy (VA) interviews, in a format similar to the input of the widely used 'InterVA' method. This package provides general model fitting and customization for 'InSilicoVA' algorithm and basic graphical visualization of the output.

Copy Link

Version

Install

install.packages('InSilicoVA')

Monthly Downloads

394

Version

1.4.0

License

GPL-2

Issues

Pull Requests

Stars

Forks

Repository

https://github.com/verbal-autopsy-software/InSilicoVA

Maintainer

Zehang Li

Last Published

September 29th, 2022

Functions in InSilicoVA (1.4.0)

Modified InSilicoVA methods with training data

Implement InSilicoVA methods with more flexible customization

Get individual COD probabilities from InSilicoVA Model Fits

Print method for summarizing InSilicoVA Model Fits

Map ICD-10 codes into the WHO 2016 cause list

plot CSMF from a "insilico" object

plot aggregated COD distribution

Message to set heap size for Java

Implement InSilicoVA methods

physician_debias

Implement physician debias algorithm

Conditional probability of InterVA4

print.insilico_summary

Print method for summarizing InSilicoVA Model Fits

summary.insilico

Summarizing InSilicoVA Model Fits

Update individual COD probabilities from InSilicoVA Model Fits

plot grouped CSMF from a "insilico" object

Conditional probability of InterVA4.03

Conditional probability table used by InterVA-4

Obtain conditional probabilities from training data

SamplePhysician

100 records of Sample debiased physician codes

Translation list of COD codes

Correspondence between InterVA causes and the physician coded cause categories

Conditional probability values used by InterVA-4

100 records of Sample Input

1000 records of Sample Input

Convergence test for fitted InSilico model

RandomPhysician

100 records of Sample Input together with two simulated physician codes