opt_int: Optimum intervention under the LAGO design

In Arhit-Chakrabarti/logisticLAGO: Learn-As-You-Go study with a single constraint and binary response

Description

The first goal of a Learn-As-You-Go (LAGO) study is to identify the optimal intervention package while minimizing the cost of the intervention package subject to the probability of a desired binary outcome being above a given threshold. The current LAGO design considers a logistic regression model with linear cost coefficients. This function takes as input the vector of linear cost coefficients, the true/estimated beta values, the lower and upper limits for the components of the intervention package, the desired outcome goal and a starting value of the intervention package for the optimization algorithm and calculates the optimal intervention package.

Usage

opt_int(
  cost,
  beta,
  lower,
  upper,
  starting.value,
  pstar,
  intercept = TRUE,
  eps = 1e-07,
  max_eval = 3000
)

Arguments

cost	a vector of linear cost coefficients
beta	a vector of true/estimated beta values
lower	a vector providing the values of lower limits for the components of the intervention package
upper	a vector providing the values of upper limits for the components of the intervention package
starting.value	a vector of starting values for the components of the intervention package
pstar	the desired outcome goal
intercept	a logical argument to include intercept in the model or not (default = TRUE)
eps	the desired level of tolerance (default = 1.0e-7)
max_eval	the maximum number of iterations to perform (default = 3000)

Value

The returned value is a named list containing

Optimum_Intervention	the optimum intervention package
Obtained_p	the outcome goal under the optimal intervention package

Examples

# Defining vector of starting values for the algorithm
x.init = c(2.5, 12.5, 7)
# Defining vector of lower limits for the components
x.l = c(1, 10, 2)
# Defining vector of upper limits for the components
x.u = c(4, 15, 15)
# Defining vector of linear cost coefficients
cost_lin = c(1, 8, 2.5)
p_bar = 0.9 # Defining the desired outcome goal
# True beta values for the study
beta = c(log(0.05), log(1.2), log(1.1), log(1.3))
## Running the LAGO optimization algorithm
opt_lago = opt_int(cost = cost_lin, beta = beta, lower = x.l,
                   upper = x.u, pstar = p_bar, starting.value = x.init)

Arhit-Chakrabarti/logisticLAGO documentation built on Dec. 17, 2021, 9:43 a.m.