generate_TGI_summary_statistics: Create a list with TGI and TGI_baseline summary statistics
In wfforrest/maeve: Modeling Accessories Enabling Visualization of Experiments

Description Usage Arguments Value Author(s) References Examples

View source: R/generate_TGI_summary_statistics.R

Create a list with TGI and TGI_baseline summary statistics

generate_TGI_summary_statistics(
  model_list,
  clean_DF,
  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 = levels(model_list$data$clean_DF_pred$study_id),
  metric = c("AUC", "ITGR", "linear"),
  progress = FALSE,
  xmin = NULL,
  xmax = NULL,
  norm_to_xrange = TRUE,
  extended_output = FALSE,
  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( ..., return_all = FALSE).
`clean_DF`	data.frame with fitted data from maeve::predict_study().
`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
`norm_to_xrange`	logical whether or not to normalize estimated effects to interval length.
`extended_output`	logical whether to return a full range of inferred interim table values.
`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.