args <- commandArgs(trailingOnly = TRUE)
library(DAISIE)
library(islandpaleoarea)
array_index <- as.numeric(args[1])
time_slice <- as.numeric(args[2])
methode <- args[3]
optimmethod <- args[4]
model <- array_index
parallel <- "local"
data_name <- data(archipelagos41_paleo)
seed <- as.integer(Sys.time()) %% 10000L * array_index
set.seed(
seed,
kind = "Mersenne-Twister",
normal.kind = "Inversion",
sample.kind = "Rejection"
)
DAISIEutils::print_metadata(
data_name = paste(data_name, model, time_slice, "prev_time", sep = "_"),
array_index = "prev_time",
model = model,
seed = seed,
methode = methode,
optimmethod = optimmethod
)
output_folder_path <- DAISIEutils::create_output_folder(
data_name = data_name,
results_dir = NULL
)
# Find previous timeslice results
prev_time_slice <- time_slice - 1
if (prev_time_slice > 0) {
files_to_read <- list.files(
path = output_folder_path,
pattern = paste0(data_name, "_", model, "_", prev_time_slice, "_"),
full.names = TRUE
)
if (length(files_to_read) <= 0) {
stop("No files found.")
}
previous_time_slice_res <- data.frame(
age = rep(NA, length(files_to_read)),
model = rep(NA, length(files_to_read)),
seed = rep(NA, length(files_to_read)),
lambda_c0 = rep(NA, length(files_to_read)),
y = rep(NA, length(files_to_read)),
mu_0 = rep(NA, length(files_to_read)),
x = rep(NA, length(files_to_read)),
K_0 = rep(NA, length(files_to_read)),
z = rep(NA, length(files_to_read)),
gamma_0 = rep(NA, length(files_to_read)),
alpha = rep(NA, length(files_to_read)),
lambda_a0 = rep(NA, length(files_to_read)),
beta = rep(NA, length(files_to_read)),
d_0 = rep(NA, length(files_to_read)),
d0_col = rep(NA, length(files_to_read)),
d0_ana = rep(NA, length(files_to_read)),
loglik = rep(NA, length(files_to_read)),
df = rep(NA, length(files_to_read))
)
}
for (i in seq_along(files_to_read)) {
message("reading file ", files_to_read[i])
input <- readRDS(files_to_read[i])
split_name <- strsplit(files_to_read[i], "_")[[1]]
previous_time_slice_res$model[i] <- as.numeric(split_name[4])
previous_time_slice_res$age[i] <- as.numeric(split_name[5])
previous_time_slice_res$seed[i] <- as.numeric(sub("*.rds.*", "\\1", split_name[6]))
previous_time_slice_res$lambda_c0[i] <- input$lambda_c0
previous_time_slice_res$y[i] <- input$y
previous_time_slice_res$mu_0[i] <- input$mu_0
previous_time_slice_res$x[i] <- input$x
previous_time_slice_res$K_0[i] <- input$K_0
previous_time_slice_res$z[i] <- input$z
previous_time_slice_res$gamma_0[i] <- input$gamma_0
previous_time_slice_res$alpha[i] <- input$alpha
previous_time_slice_res$lambda_a0[i] <- input$lambda_a0
previous_time_slice_res$beta[i] <- input$beta
previous_time_slice_res$d_0[i] <- ifelse(is.null(input$d_0), NA, input$d_0)
previous_time_slice_res$d0_col[i] <- ifelse(is.null(input$d0_col), NA, input$d0_col)
previous_time_slice_res$d0_ana[i] <- ifelse(is.null(input$d0_ana), NA, input$d0_ana)
previous_time_slice_res$loglik[i] <- input$loglik
previous_time_slice_res$df[i] <- input$df
previous_time_slice_res$conv[i] <- input$conv # TODO: Skip if not conv
}
bics <- calc_bic(
loglik = previous_time_slice_res$loglik,
df = previous_time_slice_res$df,
n = 1000
)
if (all(is.na(bics))) {
stop("Files found, but no valid previous results available.")
}
if (all(previous_time_slice_res$conv != 0)) {
stop("Files found, nothing converged.")
}
best_previous_time_slice <- previous_time_slice_res[which(bics == sort(bics)[1]), ]
message("Using parameters from preceeding time slice.")
message("Files to read were: ", files_to_read)
message("The current time slice is: ", time_slice)
message("The previous time slice is: ", time_slice - 1)
message("The previous time slice initpars are: ", paste(unlist(best_previous_time_slice[4:16]), collapse = " "))
datalist <- archipelagos41_paleo[[time_slice]]
model_args <- setup_mw_model_fixed_pars(model, best_previous_time_slice)
initparsopt <- model_args$initparsopt
idparsopt <- model_args$idparsopt
parsfix <- model_args$parsfix
idparsfix <- model_args$idparsfix
res <- model_args$res
ddmodel <- model_args$ddmodel
cpus <- model_args$cpus
tol <- model_args$tol
distance_type <- model_args$distance_type
distance_dep <- model_args$distance_dep
lik_res <- DAISIE::DAISIE_MW_ML(
datalist = datalist,
initparsopt = initparsopt,
idparsopt = idparsopt,
parsfix = parsfix,
idparsfix = idparsfix,
res = res,
ddmodel = ddmodel,
methode = methode,
cpus = cpus,
parallel = parallel,
optimmethod = optimmethod,
tol = tol,
distance_type = distance_type,
distance_dep = distance_dep
)
output_path <- file.path(
output_folder_path,
paste0(
data_name,
"_",
model,
"_",
time_slice,
"_prev_time.rds")
)
saveRDS(
lik_res,
file = output_path
)
if (!file.exists(output_path)) {
message("Error writing RDS file.")
}
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.