Skip to contents

Project a trained RCI object onto new data

Source: R/ProjectRCI.R
ProjectRCI.Rd

Apply previously learned regression-based Reliable Change Index (RCI) models to a new or expanded longitudinal dataset WITHOUT refitting models.

Usage

ProjectRCI(
  data,
  Object,
  id_var = NULL,
  Data = lifecycle::deprecated(),
  ID = lifecycle::deprecated()
)

Arguments

data

A data frame.

Object

A SciDataReportR_RCI object created with CreateRCIObject().

id_var

Optional ID column override.

Data

Deprecated (since 19.15.0). Use data instead.

ID

Deprecated (since 19.15.0). Use id_var instead.

Value

A projected SciDataReportR_RCI object.

Details

Supports both wide and long data structures and returns:

  • tidy longitudinal RCI values

  • original data merged with projected RCI outputs

  • publication-ready plots

Examples

if (FALSE) { # \dontrun{
# NOTE: This example is not run because ProjectRCI() currently errors with
# "object 'VariableTable' not found" - VariableTable is used but never
# defined in the function body (tracked bug, to be fixed separately).
rci_data <- data.frame(
  id = rep(1:30, each = 2),
  visit = rep(c("Baseline", "Followup"), 30),
  Score = round(rnorm(60, mean = 50, sd = 10), 1)
)

rci <- CreateRCIObject(
  data = rci_data,
  variables = "Score",
  DataFormat = "long",
  id_var = "id",
  VisitColumn = "visit",
  BaselineVisit = "Baseline"
)

projected <- ProjectRCI(data = rci_data, Object = rci, id_var = "id")
projected$Plots$Spaghetti$Score
} # }