Description Usage Arguments Details Value References Examples
Estimate the average causal effect (ACE) of a given treatment variable
X on a given outcome Y by adjusting for a set S of covariates (equivalenly, assuming
S blocks all backdoor paths in the causal graph that generates the observations). . Bounds are based on finding conditional instrumental variables
using the faithfulness assumption relaxed to allow for a moderate degree of unfaithfulness.
Candidate models are generated from the method described in covsearch
.
1 | bindagCausalEffectBackdoor(problem, prior_table, S)
|
problem |
a |
prior_table |
effective sample size hyperparameter of a Dirichlet prior for testing independence with contingency tables. |
S |
an array indicating the indices of the covariates used in the adjustment. |
The algorithm implemented is a naive one. If the dimensionality of S is larger than around 20, this may crash the system.
This function is assumed to be used along other methods such as covsearch
and wpp
, which generate
relativelty small covariate sets.
the estimated ACE.
Pearl, J. (2000) Causality: Models, Reasoning and Inference. Cambridge University Press.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 | ## Generate a synthetic problem
problem <- simulateWitnessModel(p = 4, q = 4, par_max = 3, M = 1000)
## Idealized case: suppose we know the true distribution,
## get "exact" ACE estimands for different adjustment sets
sol_pop <- covsearch(problem, pop_solve = TRUE)
effect_pop <- synthesizeCausalEffect(problem)
cat(sprintf(
"ACE (true) = %1.2f\nACE (adjusting for all) = %1.2f\nACE (adjusting for nothing) = %1.2f\n",
effect_pop$effect_real, effect_pop$effect_naive, effect_pop$effect_naive2))
## Perform inference and report results
covariate_hat <- covsearch(problem, cred_calc = TRUE, M = 1000)
if (length(covariate_hat$witness) > 1) {
sol_ACE <- bindagCausalEffectBackdoor(problem, prior_table = 10, S = covariate_hat$Z[[1]])
cat(sprintf("Estimated ACE = %1.2f\n", sol_ACE))
}
|
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.