Skip to contents

Use this function to return a list of tuning parameters to analyze your diagnostic test

Usage

dx(
  data,
  classlabels = c("Negative", "Positive"),
  threshold_range = NA,
  outcome_label = NA,
  pred_varname,
  true_varname,
  setthreshold = 0.5,
  poslabel = 1,
  grouping_variables = NA,
  prevalence = NA,
  citype = "exact",
  bootreps = 2000,
  bootseed = 20191015,
  doboot = FALSE,
  direction = "auto",
  ...
)

Arguments

data

A tbl.

classlabels

Labels for predicted variable. Needs to be 0, 1 order.

threshold_range

Optional. A numeric vector of thresholds to loop over.

outcome_label

Label for outcome (string)

pred_varname

Column name containing AI prediction (string)

true_varname

Column name containing AI reference standard (string)

setthreshold

A numeric value representing the threshold used to identify AI prediction

poslabel

Positive class. Variable should be coded as 0/1 with 1 being the event

grouping_variables

Character vector of variable names to be summarized by. Variables are converted to factors if not already one.

prevalence

Numeric value between 0 and 1, representing a target prevalence for additional NPV and PPV calculations.

citype

Confidence interval type.

bootreps

Number of bootstrap samples used to generate F1 score CI

bootseed

Seed value to be used when calculating bootsraped CI's

doboot

Logical. Generate bootstrap estimate of F1 confidence interval?

direction

Direction for roc comparison. See ?pROC::roc

...

currently unused