DynTxRegime (version 3.01)

plugInValue: Estimate Plug-in Value

Description

Estimate the plug-in value of a fixed treatment regime.

Usage

plugInValue(optTx1, optTx2, response, tx1, tx2)

Arguments

optTx1

Object of class "vector." First-stage treatments corresponding to the first-stage decision rule of the proposed regime.

optTx2

Object of class "vector." Second-stage treatments corresponding to the second-stage decision rule of the proposed regime.

response

Object of class "vector." Outcome of interest.

tx1

Object of class "vector." First-stage randomized treatments.

tx2

Object of class "vector." Second-stage randomized treatments.

Value

value

estimated plug-in value of the regime

fixedReg

estimated plug-in value of all possible fixed regimes

Details

The formula for the plug-in value estimate is

$$\frac{\sum_i Y_i*ind1_i*ind1_i}{\sum_i ind1_i*ind2_i }$$

where \(ind1\) and \(ind2\) are indicators that the first- and second-stage randomized treatments were consistent with the decision rules.

References

Laber, E.B., Linn, K.A., and Stefanski, L.A. (2014). Interactive Q-learning. Biometrika, 101, 831-847.

Examples

# Load and process data set 
  data(bmiData)

  # define response y to be the negative 12 month
  # change in BMI from baseline
  bmiData$y <- -100*(bmiData[,6] - bmiData[,4])/bmiData[,4]

  # generate examples of optimal treatments
  optTx1 <- 2*rbinom(210, 1, 0.4)-1
  optTx2 <- 2*rbinom(210, 1, 0.5)-1

# Plug-in Values
  tx1 <- numeric(nrow(bmiData)) + 1.0
  tx1[bmiData$A1 == "CD"] <- -1.0

  tx2 <- numeric(nrow(bmiData)) + 1.0
  tx2[bmiData$A2 == "CD"] <- -1.0

  plugInValue(optTx1 = optTx1, 
              optTx2 = optTx2, 
              response = bmiData$y,   
              tx1 = tx1,   
              tx2 = tx2)