nbc4va (version 1.2)

Bayes Classifier for Verbal Autopsy Data

Description

An implementation of the Naive Bayes Classifier (NBC) algorithm used for Verbal Autopsy (VA) built on code from Miasnikof et al (2015) .

Copy Link

Version

Install

install.packages('nbc4va')

Monthly Downloads

176

Version

1.2

License

GPL-3

Maintainer

Richard Wen

Last Published

May 10th, 2022

Functions in nbc4va (1.2)

internalRoundFixedSum

Round values to whole numbers while preserving the sum

csmf.nbc

Calculate predicted CSMFs from a NBC model

internalGetMetrics

Calculate overall performance metrics

internalCheckNBCSummary

Check arguments for summary.nbc()

internalGetCauseMetrics

Calculate performance metrics table per cause

internalSubAsRest

Substitute values in a dataframe proportionally to all other values

internalGetCSMFMaxError

Calculate CSMF maximum error

internalNBC

NBC algorithm source code

internalCheckNBC

Check arguments for nbc()

internalGetCSMFAcc

Calculate CSMF accuracy

nbc4vaGUI

Web-based graphical user interface in nbc4va

nbc4vaIO

Run nbc4va using file input and output

topCOD.nbc

Cause of death predictions from a NBC model

summary.nbc

Summarize a NBC model with metrics

print.nbc_summary

Print top predicted causes from a NBC model

nbc4vaData

Example of clean data in nbc4va

nbc4vaDataRaw

Example of unclean data in nbc4va

nbc4va

nbc4va: Bayes Classifier for Verbal Autopsy Data

nbc

Train a NBC model

plot.nbc

Bar plot of top predicted causes from a NBC model

ova2nbc

Translate open verbal autopsy arguments to train a NBC model