Pin the true APE of a DGP (inversion step)

Solves for the focal coefficient such that the data-generating process in dgp has average partial effect exactly target (in probability points): the average discrete change for a binary focal variable, the average derivative for a continuous one. Root-finding runs on the deterministic integration set. Errors if the target is infeasible – for a binary focal the APE must lie inside (-p0, 1 - p0) given baseline rate p0; for a continuous focal the APE plateaus in the coefficient, and the attainable range is reported. For DGPs with a moderator, use set_aie() instead.

Usage

set_ape(dgp, target)

Arguments

dgp

A powerape_dgp from ape_dgp() or ape_dgp_empirical(), without a moderator.

target

Assumed true APE (the planning value), e.g. 0.05 for five probability points. Always explicit – there is no default truth.

Value

The dgp with the focal coefficient and target_est filled in (and, for a binary focal, the implied P(Y = 1 | focal = 1)).

Examples

d <- ape_dgp(focal = pa_var("treat", "binary", p = 0.5), baseline = 0.30)
d <- set_ape(d, target = 0.10)
d
#> powerape DGP -- probit
#>   focal: treat (binary, p = 0.50)
#>   covariates: none
#>   baseline P(Y=1 | reference) = 0.300, nuisance signal (latent pseudo-R2) = 0.000
#>   true APE pinned at +0.1000 (beta_focal = 0.2711, implied P(Y=1 | focal=1) = 0.400)