summary.insilico: Summarizing InSilicoVA Model Fits

In InSilicoVA: Probabilistic Verbal Autopsy Coding with 'InSilicoVA' Algorithm

Summarizing InSilicoVA Model Fits

Description

This function is the summary method for class insilico.

Usage

## S3 method for class 'insilico'
summary(
  object,
  CI.csmf = 0.95,
  CI.cond = 0.95,
  file = NULL,
  top = 10,
  id = NULL,
  ...
)

Arguments

object	Fitted "insilico" object.
CI.csmf	Confidence interval for CSMF estimates.
CI.cond	Confidence interval for conditional probability estimates
file	Optional .csv file to write to. If it is specified, individual cause of death distribution will be saved to the file.
top	Number of top causes to display on screen.
id	ID of specific death to display on screen.
...	Not used.

Details

summary.insilico formats some basic information about the InSilicoVA fitted object on screen and show the several top CSMFs of user's choice. See below for more detail.

Value

id.all	all IDs of the deaths.
indiv	individual Cause of Death distribution matrix.
csmf	CSMF distribution and confidence interval for each cause.
csmf.ordered	CSMF distribution and confidence interval for each cause, ordered by mean.
condprob	Conditional probability matrix and confidence intervals.
updateCondProb	Component of "insilico" object.
keepProbbase.level	Component of "insilico" object.
datacheck	Component of "insilico" object.
Nsim	Component of "insilico" object.
thin	Component of "insilico" object.
burnin	Component of "insilico" object.
jump.scale	Component of "insilico" object.
levels.prior	Component of "insilico" object.
levels.strength	Component of "insilico" object.
trunc.min	Component of "insilico" object.
trunc.max	Component of "insilico" object.
subpop_counts	Component of "insilico" object.
showTop	Component of "insilico" object.

Author(s)

Zehang Li, Tyler McCormick, Sam Clark

Maintainer: Zehang Li <[e-mail]>

References

Tyler H. McCormick, Zehang R. Li, Clara Calvert, Amelia C. Crampin, Kathleen Kahn and Samuel J. Clark Probabilistic cause-of-death assignment using verbal autopsies, Journal of the American Statistical Association (2016), 111(515):1036-1049.

See Also

insilico, plot.insilico

Examples

## Not run: 
# load sample data together with sub-population list
data(RandomVA1)
# extract InterVA style input data
data <- RandomVA1$data
# extract sub-population information. 
# The groups are "HIV Positive", "HIV Negative" and "HIV status unknown".
subpop <- RandomVA1$subpop

# run without subpopulation
fit1<- insilico( data, subpop = NULL, 
              Nsim = 400, burnin = 200, thin = 10 , seed = 1,
              external.sep = TRUE, keepProbbase.level = TRUE)
summary(fit1)
summary(fit1, top = 10)

# save individual COD distributions to files
summary(fit1, file = "results.csv")

## End(Not run)

InSilicoVA documentation built on Sept. 29, 2022, 9:06 a.m.