Calculates the Effect Modification Shift Parameter — calc_final_effect_mod

This function uses a decision tree estimator to regress the difference in expected outcomes of an exposure onto a covariate. Both the exposure and the covariate have been determined through data-adaptive methods. If no rules are found, a median split is applied.

calc_final_effect_mod_param(
  tmle_fit_av,
  tmle_fit_at,
  exposure,
  at,
  av,
  effect_m_name,
  fold_k
)

Arguments

tmle_fit_av: TMLE results for the validation fold.
tmle_fit_at: TMLE results for the training fold.
exposure: The identified exposure variable as a character string.
at: Training dataset.
av: Validation dataset.
effect_m_name: The name of the effect modifier variable as a character string.
fold_k: The fold in which the effect modification was found.