get_what.easy_tool: An S3 Method for Extraction of Components from 'easy_tool'...

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

An S3 Method for Extraction of Components from easy_tool Objects

Description

get_what.easy_tool extracts components from easy_tool-class objects.

Usage

## S3 method for class 'easy_tool'
get_what(
  from = NULL,
  what = NULL,
  ...,
  bootreps = 4000,
  conf_level = 0.95,
  se_min = 0.8
)

Arguments

from	the easy_tool-class object from which to extract the component.
what	the (character) name of the component to extract. Valid values are ⁠"Call"⁠, ⁠"QuestionWeights"⁠, ⁠"ROCci"⁠, ⁠"ROC"⁠ and ⁠"Scores"⁠. See Details.
...	optional arguments to get_what methods.
bootreps	the number of bootstrap replications for estimation of confidence intervals for what = ⁠"ROCci"⁠. Default: 4000.
conf_level	(optional) confidence level for what = ⁠ROCci⁠.
se_min	minimum value of sensitivity printed for what = ⁠ROCci⁠. Default: 0.8.

Details

get_what is provided to enable easy extraction of components that are not provided by the plot, predict, print or summary methods.

Valid values of what are:

⁠"Call"⁠

returns the function call that created from.

⁠"QuestionWeights"⁠

returns the screening question weights, which are the re-scaled logistic-regression coefficients.

⁠ROCci⁠

returns a data frame containing sensitivities, specificities and their confidence limits, and thresholds

⁠"Scores"⁠

returns the screening scores for each subject, which are the sums of the products of the binary question responses and their QuestionWeights

⁠"ROC"⁠

returns the receiver-operating characteristic for the Scores

Value

get_what.easy_tool returns (invisibly) the object specified by what.

Examples

## Not run: 
attach(uniobj1)
tool <- easy_tool(uniobj1, max = 3, crossval = TRUE)
## Get and print sensitivities and specificities at thresholds for the
##   local maxima of the ROC curve
ROCci <- get_what(from = tool, what = "ROCci")
print(ROCci)

## End(Not run)

sgutreuter/screenr documentation built on Oct. 19, 2024, 12:49 p.m.