bootstrapBmd: Estimate lower and upper bound for model-averaged bmd using...
In alfcrisci/bmdModeling: Benchmark dose modeling

Estimate lower and upper bound for model-averaged bmd using parametric bootstrap

1
2
3

bootstrapBmd(proastData, weights, modelResults, shinyInput,
  naiveApproach = FALSE, aicNull = NA, nBootstraps = 200, seed = 1,
  showProgress = FALSE)

`proastData`	list, data in proast format as returned by f.scan()
`weights`	numeric vector, estimated weights as returned by calculateWeights()
`modelResults`	list, with results for each model, same length as weights. For each model a list with at least npar, loglik, model.ans, regr.par, CES and ces.ans; these are by default included in result from f.proast(). Eventually contains also fct1 and fct2 if factors are included for the model parameters.
`shinyInput`	list with necessary parameters used for fitted models in modelResults
`naiveApproach`	boolean, TRUE if the model-averaged BMD is estimated as the weighted average of bmd values, FALSE if the model-averaged BMD is estimated based on weighted average of response values; default value is FALSE
`aicNull`	numeric, aic value for null model as criterion for accepting bootstrap data, if NA all bootstrap data are accepted; default value is NA
`nBootstraps`	integer, the number of bootstrap data sets to generate; default value is 200
`seed`	integer, allows reproducing results; default value is 1
`showProgress`	boolean, whether progress bar should be shown in shiny application; important: only use this option when function is called from within shiny application; default value is FALSE

list with modelResults and bootstrapBmd. The modelResults contain modelResults for each bootstrap data set; bootstrapBmd is data frame with all estimated bmd values per group

alfcrisci/bmdModeling documentation built on May 28, 2019, 12:32 a.m.