Compute AIPW/doubly robust scores. — aw

Computes AIPW/doubly robust scores based on observed rewards, pulled arms, and inverse probability scores. If mu_hat is provided, compute AIPW scores, otherwise compute IPW scores.

Usage

aw_scores(yobs, ws, balwts, K, mu_hat = NULL)

Arguments

yobs: Numeric vector. Observed rewards. Must not contain NA values.
ws: Integer vector. Pulled arms. Must not contain NA values. Length must match yobs.
balwts: Numeric matrix. Inverse probability score \(1[W_t=w]/e_t(w)\) of pulling arms, shape [A, K], where A is the number of observations and K is the number of arms. Must not contain NA values.
K: Integer. Number of arms. Must be a positive integer.
mu_hat: Optional numeric matrix. Plug-in estimator of arm outcomes, shape [A, K], or NULL. Must not contain NA values if provided.

Value

Numeric matrix. AIPW scores, shape [A, K].

Examples

aw_scores(yobs = c(0.5, 1, 0, 1.5),
          ws = c(1, 2, 2, 3),
          balwts = matrix(c(0.5, 2, 1, 0.5,
                            1, 1.5, 0.5, 1.5,
                            2, 1.5, 0.5, 1),
                            ncol = 3),
          K = 3,
          mu_hat = matrix(c(0.5, 0.8, 0.6, 0.3,
                            0.9, 0.2, 0.5, 0.7,
                            0.4, 0.8, 0.2, 0.6),
                            ncol = 3))
#>      [,1] [,2] [,3]
#> [1,]  0.5 0.90  0.4
#> [2,]  0.8 1.40  0.8
#> [3,]  0.6 0.25  0.2
#> [4,]  0.3 0.70  1.5