src/install.libs.R

In BANOVA: Hierarchical Bayesian ANOVA Models

# packageStartupMessage('Compiling models which will take a while...')
# lib_dir <- file.path(R_PACKAGE_DIR, paste('libs', R_ARCH, sep = ""))
# dir.create(lib_dir, recursive = TRUE, showWarnings = FALSE)
# packageStartupMessage(paste('Writing models to:', lib_dir))
# #packageStartupMessage('No models compiled currently!')
# BANOVA.model <- function (model_name, 
#                           single_level = F){
#   if(model_name %in% c('Normal', 'Poisson', 'T', 'Bernoulli', 
#                        'Binomial', 'ordMultinomial', 'Multinomial')){
#     if(single_level){
#       name <- paste("single_",model_name, ".stan", sep = "")
#     }else{
#       name <- paste(model_name, "Normal.stan", sep = "_")
#     }
#     file_src <- file.path(R_PACKAGE_SOURCE, 'inst', 'stan', name)
#     #file_src <- system.file(name, package = 'BANOVA', mustWork = TRUE)
#     #file_src <- paste("BANOVA_v9/BANOVA_R/inst/stan/",name,sep = "")
#     model_code = readChar(file_src, nchars=1000000)
#     
#   }else{
#     stop(paste(model_name, " model is not supported currently!"))
#   }
#   sol <- list(model_code = model_code, model_name = model_name, single_level = single_level)
#   class(sol) <- 'BANOVA.model'
#   return(sol)
# }
# 
# BANOVA.build <- function(BANOVA_model){
#   if(!is(BANOVA_model, 'BANOVA.model')) stop('BANOVA_model must be a BANOVA.model object, use the BANOVA.model function to create a model first!')
#   cat('Compiling the', BANOVA_model$model_name, 'model...\n')
#   stan_c <- rstan::stanc(model_code = BANOVA_model$model_code, model_name = BANOVA_model$model_name)
#   utils::capture.output(stanmodel <- rstan::stan_model(stanc_ret = stan_c,save_dso = T), type = "output")
#   cat('Compiled successfully\n')
#   sol <- list(stanmodel = stanmodel, model_name = BANOVA_model$model_name, single_level = BANOVA_model$single_level)
#   class(sol) <- 'BANOVA.build'
#   return(sol)
# }
# 
# model_Normal_2 <- BANOVA.model('Normal')
# Normal_Normal_stanmodel <- BANOVA.build(BANOVA_model = model_Normal_2)
# 
# model_T_2 <- BANOVA.model('T', single_level = F)
# T_Normal_stanmodel <- BANOVA.build(BANOVA_model = model_T_2)
# 
# model_P_2 <- BANOVA.model('Poisson', single_level = F)
# Poisson_Normal_stanmodel <- BANOVA.build(BANOVA_model = model_P_2)
# 
# model_Bern_2 <- BANOVA.model('Bernoulli', single_level = F)
# Bernoulli_Normal_stanmodel <- BANOVA.build(BANOVA_model = model_Bern_2)
# 
# model_Bin_2 <- BANOVA.model('Binomial', single_level = F)
# Binomial_Normal_stanmodel <- BANOVA.build(BANOVA_model = model_Bin_2)
# 
# model_ordMulti_2 <- BANOVA.model('ordMultinomial', single_level = F)
# ordMultinomial_Normal_stanmodel <- BANOVA.build(BANOVA_model = model_ordMulti_2 )
# 
# model_Multi_2 <- BANOVA.model('Multinomial', single_level = F)
# Multinomial_Normal_stanmodel <- BANOVA.build(BANOVA_model = model_Multi_2)
# 
# model_l2 <- file.path(lib_dir, 'BANOVA.RData')
# save(Normal_Normal_stanmodel, Binomial_Normal_stanmodel, Bernoulli_Normal_stanmodel,
#      T_Normal_stanmodel, Poisson_Normal_stanmodel, ordMultinomial_Normal_stanmodel,
#      Multinomial_Normal_stanmodel, file = model_l2)
# 
# model_Normal_1 <- BANOVA.model('Normal', single_level = T)
# Normal_Normal_stanmodel_1 <- BANOVA.build(BANOVA_model = model_Normal_1)
# 
# model_T_1 <- BANOVA.model('T', single_level = T)
# T_Normal_stanmodel_1 <- BANOVA.build(BANOVA_model = model_T_1)
# 
# model_P_1 <- BANOVA.model('Poisson', single_level = T)
# Poisson_Normal_stanmodel_1 <- BANOVA.build(BANOVA_model = model_P_1)
# 
# model_Bern_1 <- BANOVA.model('Bernoulli', single_level = T)
# Bernoulli_Normal_stanmodel_1 <- BANOVA.build(BANOVA_model = model_Bern_1)
# 
# model_Bin_1 <- BANOVA.model('Binomial', single_level = T)
# Binomial_Normal_stanmodel_1 <- BANOVA.build(BANOVA_model = model_Bin_1)
# 
# model_ordMulti_1 <- BANOVA.model('ordMultinomial', single_level = T)
# ordMultinomial_Normal_stanmodel_1 <- BANOVA.build(BANOVA_model = model_ordMulti_1 )
# 
# model_Multi_1 <- BANOVA.model('Multinomial', single_level = T)
# Multinomial_Normal_stanmodel_1 <- BANOVA.build(BANOVA_model = model_Multi_1)
# 
# model_l1 <- file.path(lib_dir, 'single_BANOVA.RData')
# save(Normal_Normal_stanmodel_1, Binomial_Normal_stanmodel_1, Bernoulli_Normal_stanmodel_1,
#      T_Normal_stanmodel_1, Poisson_Normal_stanmodel_1, ordMultinomial_Normal_stanmodel_1,
#      Multinomial_Normal_stanmodel_1, file = model_l1)
# 
# packageStartupMessage('Models compiled successfully!')