qLearn: Q-learning

Description

As a single call, regresses the response on covariates and treatment histories to estimate the optimal treatment decision rule. If called repeatedly, once for each treatment decision point, each call is a step of the Q-Learning algorithm.

Usage

qLearn(..., moMain, moCont, data, response, txName, fSet, iter = 0L, verbose = TRUE)

Arguments

…

ignored. Included to require named input.

moMain

A single object of class "modelObj" or an object of class "list" containing objects of class "modelObjSubset." This object specifies the model(s) of the main effects component of the outcome regression and the R methods to be used to obtain parameter estimates and predictions. The method chosen to obtain predictions must return the prediction on the scale of the response variable.

moCont

A single object of class "modelObj" or an object of class "list" containing objects of class "modelObjSubset." This object specifies the models of the contrasts component of the outcome regression and the R methods to be used to obtain parameter estimates and predictions. The method chosen to obtain predictions must return the prediction on the scale of the response variable.

data

An object of class "data.frame." The covariates and treatment histories.

response

An object of class "vector" or "QLearn." If performing a single decision point analysis or the first step of the Q-Learning algorithm, i.e., the final-stage regression, response is a vector of the outcome of interest. For multiple decision point analysis, this is the value object returned by the previous call to qLearn().

txName

An object of class "character." The column header of the stage treatment variable as given in input data.

fSet

An object of class "function" or NULL. See fSet for further information.

iter

An object of class "numeric." If >0, the iterative method will be used to obtain parameter estimates in the outcome regression step. See iter for further information.

verbose

An object of class "logical." If FALSE, screen prints will be suppressed.

Value

Returns an object of class "QLearn" that inherits directly from class "DynTxRegime."

Methods

coef: signature(object = "QLearn"): Retrieve parameter estimates for all regression steps.
DTRstep: signature(object = "QLearn"): Retrieve description of method used to create object.
estimator: signature(x = "QLearn"): Retrieve the estimated value of the estimated optimal regime for the training data set.
fitObject: signature(object = "QLearn"): Retrieve value object returned by regression methods.
optTx: signature(x = "QLearn", newdata = "missing"): Retrieve the estimated optimal treatment regime for training data set.
optTx: signature(x = "QLearn", newdata = "data.frame"): Estimate the optimal treatment regime for newdata.
outcome: signature(x = "QLearn"): Retrieve value object returned by outcome regression methods.
plot: signature(x = "QLearn"): Generate plots for regression analyses.
print: signature(object = "QLearn"): Print main results of analysis.
show: signature(object = "QLearn"): Show main results of analysis.
summary: signature(object = "QLearn"): Retrieve summary information from regression analyses.

Examples

Run this code

# NOT RUN {
# 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]

# Second-stage regression
  # Create modeling object for main effect component
  moMain <- buildModelObj(model = ~ gender + parentBMI + month4BMI,
                          solver.method='lm')

  # Create modeling object for contrast component
  moCont <- buildModelObj(model = ~ parentBMI + month4BMI,
                          solver.method='lm')

  fitQ2 <- qLearn(moMain = moMain, 
                  moCont = moCont, 
                  data = bmiData,  
                  response = bmiData$y, 
                  txName = "A2", 
                  iter = 0L)

##Available methods

  # Coefficients of the propensity score regression
  coef(fitQ2)

  # Description of method used to obtain object
  DTRstep(fitQ2)

  # Estimated value of the optimal treatment regime for training set
  estimator(fitQ2)

  # Value object returned by propensity score regression method
  fitObject(fitQ2)

  # Estimated optimal treatment for training data
  optTx(fitQ2)

  # Estimated optimal treatment for new data
  optTx(fitQ2, bmiData)

  # Value object returned by outcome regression method
  outcome(fitQ2)

  # Plots if defined by propensity regression method
  dev.new()
  par(mfrow = c(2,4))

  plot(fitQ2)
  plot(fitQ2, suppress = TRUE)

  # Show main results of method
  show(fitQ2)

  # Show summary results of method
  summary(fitQ2)
 

# First-stage regression

  # Create modeling object for main effect component
  moMain <- buildModelObj(model = ~ gender + race + parentBMI + baselineBMI,
                          solver.method='lm')

  # Create modeling object for contrast component
  moCont <- buildModelObj(model = ~ gender + parentBMI,
                          solver.method='lm')

  fitQ1 <- qLearn(moMain = moMain, 
                  moCont = moCont,  
                  response = fitQ2, 
                  data = bmiData,  
                  txName = "A1",  
                  iter = 100L)

# }