generate_summary_table: Create a summary table making Dunnett comparisons.
In wfforrest/maeve: Modeling Accessories Enabling Visualization of Experiments

Description Usage Arguments Value Author(s) References Examples

This is included principally for its options to generate legacy statistics, and is not recommended for general use.

generate_summary_table(
  model_list,
  clean_DF,
  reference_Dunnett = maeve_options("reference_Dunnett"),
  full_inverse_function = switch(maeve_options("inv_func_char"), Identity = function(x)
    {     x - maeve_options("add_to_endpoint") }, function(x) {    
    (get(maeve_options("inv_func_char")))(x) - maeve_options("add_to_endpoint") }),
  study_ID_value = "",
  metric = c("AUC", "ITGR", "linear", "AUC_pwl", "ITGR_pwl", "AUC_poly", "ITGR_poly"),
  progress = maeve_options("progress"),
  xmin = NULL,
  xmax = NULL,
  xrange_norm_method = maeve_options("xrange_norm_method"),
  extended_output = FALSE,
  conf_int = c("none", "TGI", "Effect_Duration")[1],
  EOS_CR_minval = maeve_options("EOS_CR_minval"),
  PR_threshold = maeve_options("PR_threshold"),
  CI_lower_quantile = 0.025,
  CI_upper_quantile = 0.975,
  x_time_spacing = 0.25,
  x_min_count_distinct = 10,
  K_boot = 1000
)

`model_list`	R list with model objects from maeve::model_study().
`clean_DF`	data.frame with fitted data from maeve::model_study(), e.g., "model_list$data$clean_DF_pred".
`reference_Dunnett`	character string with exact match to an element of grpnames. The exact match will be the reference group in Dunnett contrasts.
`full_inverse_function`	R function that inverts the response analyzed back to its original scale.
`study_ID_value`	character string with five digit DIVOS study ID, OR a DivoStudy object.
`metric`	character vector specifying what metric(s) to use in longitudinal modeling.
`progress`	logical determining whether to send interim progress messages.
`xmin`	numeric lower bound of definite integral
`xmax`	numeric upper bound of definite integral
`xrange_norm_method`	character string determining how the xrange normalization within leia() should be done. Must be one of "none", "xrange", "slope_equivalent".
`extended_output`	logical whether to return a full range of inferred interim table values.
`conf_int`	character vector designating for which quantities a parametric bootstrap should be run to get confidence intervals. Allowed options are "none" (the default", "TGI", "Effect_Duration", or c("TGI", "Effect_Duration").
`EOS_CR_minval`	numeric passed to internal function to decide what constitutes a "complete response" for the the End-Of-Study Complete Response (EOS_CR) value on the original, untransformed response scale.
`PR_threshold`	numeric passed to internal function to decide what constitutes a "partial response" for the PR value. Fractional reduction is computed by-ID, from an ID-specific baseline value on the original, untransformed scale.
`CI_lower_quantile`	numeric value in the range (0,1) denoting the lower quanitle for confidence intervals.
`CI_upper_quantile`	numeric value in the range (0,1) denoting the upper quanitle for confidence intervals.
`x_time_spacing`	spacing of time points across which to evaluate the splines. If time values are recorded as integers, then the default should be sufficient.
`x_min_count_distinct`	numeric integer. If the number of distinct x-values at which splines are evaluated is less than this, a warning is issued.
`K_boot`	integer number of bootstrap samples to draw.