ntpp.lasso_screenr: An S3 Method to Compute the Ratio of Total Tests to Positive...

In sgutreuter/screenr: Construction of Binary Test-Screening Rules

An S3 Method to Compute the Ratio of Total Tests to Positive Results

Description

ntpp.lasso_screenr computes the ratio of the total number of tests performed per positive test result and the anticipated proportion of the untested (those screened out of testing) who would actually test positive.

Usage

## S3 method for class 'lasso_screenr'
ntpp(
  object,
  ...,
  model = c("minAIC", "minBIC"),
  type = c("cvResults", "isResults"),
  prev = NULL
)

Arguments

object	a lasso_screenr-class object produced by lasso_screenr.
...	optional arguments to ntpp methods.
model	(character) select the model which produced the minimum AIC ("minAIC", the default) or minimum BIC ("minBIC").
type	(character) one of "cvResults" (the default) or "isResults" to specify k-fold cross-validated or in-sample receiver-operating characteristics, respectively.
prev	an optional positive proportion for the test outcome; if missing the test positivity is obtained from object.

Details

The anticipated number of tests required to detect a single positive nntp is given by

nntp = (SeP + (1 - Sp)(1 - P)) / SeP

where Se is sensitivity, P is prevalence and Sp is specificity. The anticipated positivity among those screened out is given by

Puntested = ((1 - Se)P) / ((1 - Se)P + Sp (1 - P))

Value

ntpp.lasso_screenr returns a data frame containing the following columns:

sensitivity

The sensitivity (proportion) of the screener.

specificity

The specificity (proportion) of the screener.

ntpp

the number of tests required to discover a single positive test result.

prev_untested

The antipated proportion who would positive among those who are screened out of testing.

Examples

attach(uniobj1)
ntpp(uniobj1)

sgutreuter/screenr documentation built on Nov. 20, 2022, 2:41 a.m.