|
ci()
|
Compute confidence intervals for drtmle and adaptive_iptw@ |
|
ci(<intermed>)
|
Confidence intervals for interventional mediation effects |
|
estimate_G()
|
estimateG |
|
estimate_Q_M()
|
estimate_Q_M |
|
estimate_Qbar()
|
estimate_Qbar |
|
evaluate_direct_effect()
|
Helper function to evaluate the direct effect |
|
evaluate_eif_direct()
|
Function to evaluate the canonical gradient of
the direct effect at a given set of estimated nuisance parameters
and for each observation. |
|
evaluate_eif_indirect_M1()
|
Function to evaluate the canonical gradient of
the indirect effect through M1 at a given set of estimated nuisance parameters
and for each observation. |
|
evaluate_eif_indirect_M2()
|
Function to evaluate the canonical gradient of
the indirect effect through M2 at a given set of estimated nuisance parameters
and for each observation. |
|
evaluate_eif_total()
|
Function to evaluate the canonical gradient of
the total effect at a given set of estimated nuisance parameters
and for each observation. |
|
evaluate_total_effect()
|
Helper function to evaluate the total effect |
|
extract_Qbar_obs()
|
Helper function for extracting Qbar at the observed values of confounders and
mediators under a or a_star |
|
extract_joint()
|
Helper function for extracting joint distributions at particular values of
the mediator. |
|
extract_marginal()
|
Helper function for extracting marginal distributions at particular values of
the mediator. |
|
format_long_hazards()
|
Generate long format hazards data for conditional density estimation |
|
get_Qbarbar()
|
Helper function to marginalize outcome regression over mediator
distributions. Used to get each of the relevant nuisance parameters needed
to evaluate the total, direct, and indirect effects. |
|
intermed()
|
AIPTW and TMLE estimates of interventional mediation effects |
|
map_hazard_to_density()
|
Map a predicted hazard to a predicted density for a single observation |
|
predict_density()
|
Helper function to predict from a super learner fit on the hazard scale
to obtain a prediction on the density scale |
|
print(<intermed>)
|
Print the AIPTW results |
|
reduce_merge()
|
Helper to merge covariates in targeting |
|
reorder_list()
|
Helper function to reorder lists according to cvFolds |
|
target_Qbar()
|
Target the outcome regression |
|
target_Qbarbar()
|
Target the iterated regressions |
|
target_Qbarbar_M1_star_times_M2_star_a()
|
Function for targeting Qbarbar_M1_star_times_M2_star_a |
|
target_Qbarbar_M1_times_M2_a()
|
Function for targeting Qbarbar_M1_times_M2_a |
|
target_Qbarbar_M1_times_M2_star_a()
|
Function for targeting Qbarbar_M1_times_M2_star_a |
|
target_conditional_direct_effect()
|
The naming convention is a bit different here, because we're actually
able to go after the conditional direct effect, well, directly. That is,
we can define a loss function whose minimizer defines the conditional mean of the
difference between Qbar_a
and Qbar_a_star with respect to the joint distribution of M1 and M2 given C and A = a_star.
We can then define a submodel through this conditional mean difference (which is
exactly the conditional direct effect) and target this quantity directly. |
|
target_conditional_total_effect()
|
The naming convention is a bit different here, because we're actually
able to go after the conditional total effect directly. That is,
we can define a loss function whose minimizer defines the conditional mean of the
difference between Qbar_a
and Qbar_a_star with respect to the joint distribution of M1 and M2 given C and A.
We can then define a submodel through this conditional mean difference (which is
exactly the conditional total effect) and target this quantity directly. |