est_psi_prob: Tests if A is used in the regression function, i.e. if...
In kararudolph/transport: Transported Causal Inference With Targeted Learning

Description Usage Author(s) References

View source: R/est_psi_prob.R

Tests if A is used in the regression function, i.e. if E[Y|A,W] = E[Y|W] almost surely, or equivalently (under positivity) if E[Y|A=1,W] = E[Y|A=0,W] almost surely Returns an estimate of Psi, and a p-value If est.g is false, then just uses predefined function g0 NOTE: Theory all uses bounded Y, so if Y is not bounded then at least try to make sure most of its mass falls in the interval [-1,1] — can do this by scaling by a constant

1
2
3

est_psi_prob(W, A, Y, W.train = NULL, A.train = NULL, Y.train = NULL,
  sig.meth = "var", est.g = TRUE, g0 = NULL, SL.library = c("SL.rpart",
  "SL.glm", "SL.earth", "SL.nnet", "SL.gam"))