R/clv_template_controlflow_predict.R

In bachmannpatrick/CLVTools: Tools for Customer Lifetime Value Estimation

clv.template.controlflow.predict <- function(clv.fitted, verbose, user.newdata, ...){

  # Check if can predict -----------------------------------------------------------------------------------------
  # Cannot predict if there are any NAs in any of the prediction.params
  clv.controlflow.check.prediction.params(clv.fitted = clv.fitted)

  # Process Newdata -------------------------------------------------------------------------------------
  # Because many of the following steps refer to the data stored in the fitted model,
  #   it first is replaced with newdata before any other steps are done
  if(!is.null(user.newdata)){
    # check newdata
    clv.controlflow.check.newdata(clv.fitted = clv.fitted, user.newdata = user.newdata, ...)

    # Replace data in model with newdata
    #   Deep copy to not change user input
    clv.fitted@clv.data <- copy(user.newdata)

    # Do model dependent steps of adding newdata
    clv.fitted <- clv.model.process.newdata(clv.model = clv.fitted@clv.model, clv.fitted=clv.fitted, verbose=verbose)
  }


  # Input checks ----------------------------------------------------------------------------------------
  #   Only after newdata replaced clv.data stored in clv.fitted because inputchecks use clv.fitted@clv.data
  clv.controlflow.predict.check.inputs(clv.fitted=clv.fitted, verbose=verbose, ...)

  # Prediction result table -----------------------------------------------------------------------------
  dt.predictions <- clv.controlflow.predict.build.result.table(clv.fitted=clv.fitted, verbose=verbose, ...)

  # Model prediction ------------------------------------------------------------------------------------
  dt.predictions <- clv.model.predict(clv.model = clv.fitted@clv.model, clv.fitted = clv.fitted,
                                      dt.predictions = dt.predictions, verbose = verbose, ...)
  setkeyv(dt.predictions, "Id")

  # Actuals ---------------------------------------------------------------------------------------------
  has.actuals    <- clv.controlflow.predict.get.has.actuals(clv.fitted, dt.predictions = dt.predictions)
  dt.predictions <- clv.controlflow.predict.add.actuals(clv.fitted = clv.fitted, dt.predictions = dt.predictions,
                                                        has.actuals = has.actuals, verbose = verbose, ...)

  # post.process / add any additional steps -------------------------------------------------------------
  # set col order etc
  dt.predictions <- clv.controlflow.predict.post.process.prediction.table(clv.fitted = clv.fitted,
                                                                          has.actuals = has.actuals,
                                                                          dt.predictions = dt.predictions,
                                                                          verbose = verbose, ...)


  # data.table does not print when returned because it is returned directly after last [:=]
  # " if a := is used inside a function with no DT[] before the end of the function, then the next
  #   time DT or print(DT) is typed at the prompt, nothing will be printed. A repeated DT or print(DT)
  #   will print. To avoid this: include a DT[] after the last := in your function."
  dt.predictions[]


  return(dt.predictions)
}