Assemble LMTP learner diagnostics for reporting — margot_report_lmtp

Bundles the learner heatmap, concise narrative, and a compact summary table of dominant learners for a selected outcome and set of shifts. Designed to keep the visual and textual learner diagnostics aligned in Quarto and manuscript workflows.

Usage

margot_report_lmtp_learners(
  x,
  outcome = NULL,
  shifts = NULL,
  label_mapping = NULL,
  waves = NULL,
  remove_waves = NULL,
  title = NULL,
  component = c("both", "outcome", "treatment"),
  digits = 1,
  include_plot = TRUE,
  plot_args = list(),
  interpret_args = list()
)

Arguments

x: Result of `margot_lmtp()` (with `$models`) or another object accepted by the learner helpers.
outcome: Optional character outcome name. When `NULL`, the first stored outcome is used.
shifts: Optional character vector of shifts to include (full or cleaned names). If `NULL`, all available shifts are used.
label_mapping: Optional label map passed through to downstream helpers.
waves: Optional integer vector selecting waves.
remove_waves: Optional integer vector of waves to drop after subsetting.
title: Optional plot title passed to [margot_plot_lmtp_learners()].
component: Which nuisance models to include: `"both"` (default), `"outcome"` (only `m`), or `"treatment"` (only `r`).
digits: Integer rounding applied to percentage columns in the summary table and to the narrative produced by [margot_interpret_lmtp_learners()].
include_plot: Logical; if `TRUE`, returns a `ggplot2` object from [margot_plot_lmtp_learners()].
plot_args: Optional named list overriding defaults passed to [margot_plot_lmtp_learners()].
interpret_args: Optional named list overriding defaults passed to [margot_interpret_lmtp_learners()] (for example, `digits = 1`).

Value

A named list with elements: - `summary_table`: compact dominant-learner table by shift, component, and wave. - `learner_data`: tidy learner-weight data from `summarise_lmtp_learners()`. - `plot`: learner heatmap (or `NULL` when `include_plot = FALSE`). - `narrative`: structured list from `margot_interpret_lmtp_learners(return = "list")`. - `method_statement`: single character string describing the learner diagnostics. - `metadata`: list of context (outcome, shifts, waves, component).