baker (version 1.0.4)

assign_model: Interpret the specified model structure

Description

assign_model translates options specified by a user (e.g., in model_options) into information that can be understood by baker.

Usage

assign_model(model_options, data_nplcm, silent = TRUE)

Value

A list of model specifications:

• num_slice A vector counting the No. of measurement slices for each level of measurement quality (e.g., MBS, MSS, MGS representing Bronze-Standard Measurements - case-control, Silver-Standard Measurements and Gold-Standard Measurements - case-only);

• nested Local dependence specification for modeling bronze-standard data. TRUE for nested models (conditional dependence given disease class); FALSE for non-nested models (conditional independence given disease class). One for each BrS slice.

• regression

  • do_reg_Eti TRUE for doing etiology regression. It means let the etiology fractions vary with explanatory variables. FALSE otherwise;

  • do_reg_FPR A vector whose names represent the slices of bronze-standard data. For each slice of BrS measurements, TRUE does false positive rate regression. It means the false positive rates, estimatable from controls, can vary with covariates; FALSE otherwise.

  • is_discrete_predictor A list of names "Eti", and the names for every slice of bronze-standard data. TRUE if all predictors are discrete; FALSE otherwise.

Arguments

model_options

See nplcm() function.

data_nplcm

Data. See nplcm() function for data structure.

silent

Default is TRUE for no messages; FALSE otherwise.

Details

assign_model will be modified to check if data are conformable to specified model.

Examples

cause_list <- c(LETTERS[1:6]) 
J.BrS <- 6
model_options_no_reg <- list(
likelihood   = list(
  cause_list = cause_list,
  k_subclass = 2,
  Eti_formula = ~-1, 
  # no covariate for the etiology regression
  FPR_formula = list(
    MBS1 =   ~-1)    
    # no covariate for the subclass weight regression
),
use_measurements = c("BrS"), 
# use bronze-standard data only for model estimation.
prior= list(
  Eti_prior = overall_uniform(1,cause_list), 
  # Dirichlet(1,...,1) prior for the etiology.
  TPR_prior  = list(BrS = list(
    info  = "informative", # informative prior for TPRs
    input = "match_range", 
    # specify the informative prior for TPRs by specifying a plausible range.
    val = list(MBS1 = list(up =  list(rep(0.99,J.BrS)), 
    # upper ranges: matched to 97.5% quantile of a Beta prior
                           low = list(rep(0.55,J.BrS))))
                           # lower ranges: matched to 2.5% quantile of a Beta prior
  )
  )
)
)     
data("data_nplcm_noreg")

assign_model(model_options_no_reg,data_nplcm_noreg)