Newer
Older
#'Save CSTools objects of class 's2dv_cube' containing experiments or observed
#'data in NetCDF format
#'
#'@author Perez-Zanon Nuria, \email{nuria.perez@bsc.es}
#'
#'@description This function allows to divide and save a object of class
#''s2dv_cube' into a NetCDF file, allowing to reload the saved data using
#'\code{CST_Load} function.
#'@param data An object of class \code{s2dv_cube}.
#'@param destination A character string containing the directory name in which
#' to save the data. NetCDF file for each starting date are saved into the
#' folder tree: \cr
#' destination/experiment/variable/. By default the function
#' creates and saves the data into the folder "CST_Data" in the working
#' directory.
#'@param extra_string A character string to be include as part of the file name,
#' for instance, to identify member or realization. It would be added to the
#' file name between underscore characters.
#'@param single_file A logical value indicating if all object is saved in a
#' unique file (TRUE) or in separated directories (FALSE). When it is FALSE,
#' the array is separated for Datasets, variable and start date.
#'@seealso \code{\link{CST_Load}}, \code{\link{as.s2dv_cube}} and \code{\link{s2dv_cube}}
#'data <- lonlat_temp$exp
#'CST_SaveExp(data = data, destination = destination)
#'}
#'
#'@import ncdf4
#'@importFrom s2dv Reorder InsertDim
#'@importFrom ClimProjDiags Subset
#'@import multiApply
CST_SaveExp <- function(data, destination = "./", sdate_dim = 'sdate',
ftime_dim = 'time', dat_dim = 'dataset',
var_dim = 'var', memb_dim = 'member',
single_file = TRUE, extra_string = NULL, use_load = FALSE) {
# Check 's2dv_cube'
if (!inherits(data, 's2dv_cube')) {
stop("Parameter 'data' must be of the class 's2dv_cube', ",
"as output by CSTools::CST_Load.")
}
# Check object structure
if (!all(c('data', 'attrs') %in% names(data))) {
stop("Parameter 'data' must have at least 'data' and 'attrs' elements ",
"within the 's2dv_cube' structure.")
}
if (!inherits(data$attrs, 'list')) {
stop("Level 'attrs' must be a list with at least 'Dates' element.")
}
if (!all(c('coords') %in% names(data))) {
warning("Element 'coords' not found. No coordinates will be used.")
}
if (!is.character(sdate_dim)) {
stop("Parameter 'sdate_dim' must be a character string.")
warning("Parameter 'sdate_dim' has length greater than 1 and ",
"only the first element will be used.")
}
# metadata
if (is.null(data$attrs$Variable$metadata)) {
warning("No metadata found in element Variable from attrs.")
} else {
if (!inherits(data$attrs$Variable$metadata, 'list')) {
stop("Element metadata from Variable element in attrs must be a list.")
}
if (!any(names(data$attrs$Variable$metadata) %in% names(data$coords))) {
warning("Metadata is not found for any coordinate.")
} else if (!any(names(data$attrs$Variable$metadata) %in%
data$attrs$Variable$varName)) {
warning("Metadata is not found for any variable.")
}
}
if (is.null(data$attrs$Dates)) {
stop("Element 'Dates' from 'attrs' level cannot be NULL.")
}
if (is.null(dim(data$attrs$Dates))) {
stop("Element 'Dates' from 'attrs' level must have time dimensions.")
}
SaveExp(data = data$data,
destination = destination,
Dates = data$attrs$Dates,
Datasets = data$attrs$Datasets,
startdates = data$coords[[sdate_dim]],
dat_dim = dat_dim, sdate_dim = sdate_dim,
ftime_dim = ftime_dim, var_dim = var_dim,
memb_dim = memb_dim,
#'Save an experiment in a format compatible with CST_Load
#'@description This function is created for compatibility with CST_Load/Load for
#'saving post-processed datasets such as those calibrated of downscaled with
#'CSTools functions
#'
#'@author Perez-Zanon Nuria, \email{nuria.perez@bsc.es}
#'
#'@param data An multi-dimensional array with named dimensions (longitude,
#' latitude, time, member, sdate).
#'@param destination A character string indicating the path where to store the
#' NetCDF files.
#'@param coords A named list with elements of the coordinates corresponding to
#' the dimensions of the data parameter. The names and length of each element
#' must correspond to the names of the dimensions. If any coordinate is not
#' provided, it is set as an index vector with the values from 1 to the length
#' of the corresponding dimension.
#'@param startdates A vector of dates indicating the initialization date of each
#' simulations.
#'@param Dates A matrix of dates with the corresponding sdate and forecast time
#' dimension.
#'@param Datasets A vector of character string indicating the names of the
#'@param varname A character string indicating the name of the variable to be
#'@param metadata A named list where each element is a variable containing the
#' corresponding information. The information must be contained in a list of
#' lists for each variable.
#'@param single_file A logical value indicating if all object is saved in a
#' unique file (TRUE) or in separated directories (FALSE). When it is FALSE,
#' the array is separated for Datasets, variable and start date.
#'@param extra_string A character string to be include as part of the file name,
#' for instance, to identify member or realization.
#'@return The function creates as many files as sdates per Datasets. Each file
#'could contain multiple members. It would be added to the file name between
#'underscore characters. The path will be created with the name of the variable
#'and each Datasets.
#'data <- lonlat_temp$exp$data
#'lon <- lonlat_temp$exp$coords$lon
#'lat <- lonlat_temp$exp$coords$lat
#'coords <- list(lon = lon, lat = lat)
#'Datasets <- lonlat_temp$exp$attrs$Datasets
#'varname <- 'tas'
#'Dates <- lonlat_temp$exp$attrs$Dates
#'metadata <- lonlat_temp$exp$attrs$Variable$metadata
#'SaveExp(data = data, destination = destination, coords = coords,
#' Datasets = Datasets, varname = varname, Dates = Dates,
#' metadata = metadata, single_file = TRUE)
#'@import ncdf4
#'@importFrom s2dv Reorder InsertDim
#'@import multiApply
#'@importFrom ClimProjDiags Subset
SaveExp <- function(data, destination = "./", Dates = NULL, coords = NULL,
varname = NULL, metadata = NULL, Datasets = NULL,
startdates = NULL, dat_dim = 'dataset', sdate_dim = 'sdate',
ftime_dim = 'time', var_dim = 'var', memb_dim = 'member',
single_file = TRUE, extra_string = NULL, use_load = FALSE) {
# data
if (is.null(data)) {
stop("Parameter 'data' cannot be NULL.")
}
dimnames <- names(dim(data))
if (is.null(dimnames)) {
stop("Parameter 'data' must be an array with named dimensions.")
}
# destination
if (!is.character(destination) | length(destination) > 1) {
stop("Parameter 'destination' must be a character string of one element ",
"indicating the name of the file (including the folder if needed) ",
"where the data will be saved.")
}
# Dates
if (!is.null(Dates)) {
if (!inherits(Dates, "POSIXct") & !inherits(Dates, "Date")) {
stop("Parameter 'Dates' must be of 'POSIXct' or 'Dates' class.")
}
if (is.null(dim(Dates))) {
stop("Parameter 'Dates' must have dimension names.")
}
if (!is.null(coords)) {
if (!all(names(coords) %in% dimnames)) {
coords <- coords[-which(!names(coords) %in% dimnames)]
}
for (i_coord in dimnames) {
if (i_coord %in% names(coords)) {
if (length(coords[[i_coord]]) != dim(data)[i_coord]) {
warning(paste0("Coordinate '", i_coord, "' has different lenght as ",
"its dimension and it will not be used."))
coords[[i_coord]] <- 1:dim(data)[i_coord]
}
} else {
warning(paste0("Coordinate '", i_coord, "' is not provided ",
"and it will be set as index in element coords."))
coords[[i_coord]] <- 1:dim(data)[i_coord]
}
}
} else {
# varname
if (is.null(varname)) {
warning("Parameter 'varname' is NULL. It will be assigned to 'X'.")
varname <- 'X'
} else if (length(varname) > 1) {
multiple_vars <- TRUE
} else {
multiple_vars <- FALSE
}
if (!all(sapply(varname, is.character))) {
stop("Parameter 'varname' must be a character string with the ",
"variable names.")
}
# metadata
if (is.null(metadata)) {
warning("Parameter 'metadata' is not provided so the metadata saved ",
"will be incomplete.")
}
# Spatial coordinates
if (!any(dimnames %in% .KnownLonNames()) |
!any(dimnames %in% .KnownLatNames())) {
warning("Spatial coordinate names do not match any of the names accepted by ",
"the package.")
lon_dim <- NULL
lat_dim <- NULL
} else {
lon_dim <- dimnames[which(dimnames %in% .KnownLonNames())]
lat_dim <- dimnames[which(dimnames %in% .KnownLatNames())]
if (length(lon_dim) > 1) {
warning("Found more than one longitudinal dimension. Only the first one ",
"will be used.")
lon_dim <- lon_dim[1]
}
if (length(lat_dim) > 1) {
warning("Found more than one latitudinal dimension. Only the first one ",
"will be used.")
lat_dim <- lat_dim[1]
}
if (!is.null(ftime_dim)) {
if (!is.character(ftime_dim)) {
stop("Parameter 'ftime_dim' must be a character string.")
}
if (!all(ftime_dim %in% dimnames)) {
stop("Parameter 'ftime_dim' is not found in 'data' dimension.")
}
if (length(ftime_dim) > 1) {
warning("Parameter 'ftime_dim' has length greater than 1 and ",
"only the first element will be used.")
ftime_dim <- ftime_dim[1]
}
if (!is.null(sdate_dim)) {
if (!is.character(sdate_dim)) {
stop("Parameter 'sdate_dim' must be a character string.")
}
if (length(sdate_dim) > 1) {
warning("Parameter 'sdate_dim' has length greater than 1 and ",
"only the first element will be used.")
sdate_dim <- sdate_dim[1]
}
if (!all(sdate_dim %in% dimnames)) {
stop("Parameter 'sdate_dim' is not found in 'data' dimension.")
}
}
# memb_dim
if (!is.null(memb_dim)) {
if (!is.character(memb_dim)) {
stop("Parameter 'memb_dim' must be a character string.")
}
if (!all(memb_dim %in% dimnames)) {
stop("Parameter 'memb_dim' is not found in 'data' dimension. Set it ",
"as NULL if there is no member dimension.")
}
if (use_load) {
memb_pos <- which(names(dim(data)) == memb_dim)
names(dim(data))[memb_pos] <- 'ensemble'
dimnames <- names(dim(data))
memb_dim <- 'ensemble'
}
}
# dat_dim
if (!is.null(dat_dim)) {
if (!is.character(dat_dim)) {
stop("Parameter 'dat_dim' must be a character string.")
}
if (!all(dat_dim %in% dimnames)) {
stop("Parameter 'dat_dim' is not found in 'data' dimension. Set it ",
"as NULL if there is no Datasets dimension.")
}
if (length(dat_dim) > 1) {
warning("Parameter 'dat_dim' has length greater than 1 and ",
"only the first element will be used.")
dat_dim <- dat_dim[1]
}
if (!is.null(var_dim)) {
if (!is.character(var_dim)) {
stop("Parameter 'var_dim' must be a character string.")
}
if (!all(var_dim %in% dimnames)) {
stop("Parameter 'var_dim' is not found in 'data' dimension. Set it ",
"as NULL if there is no variable dimension.")
}
if (length(var_dim) > 1) {
warning("Parameter 'var_dim' has length greater than 1 and ",
"only the first element will be used.")
var_dim <- var_dim[1]
}
n_vars <- 1
}
# minimum dimensions
if (all(dimnames %in% c(var_dim, dat_dim))) {
if (!single_file) {
warning("Parameter data has only ",
paste(c(var_dim, dat_dim), collapse = ' and '), " dimensions ",
"and it cannot be splitted in multiple files. All data will ",
"be saved in a single file.")
single_file <- TRUE
}
if (!is.null(Dates)) {
if (all(names(dim(Dates)) == c(ftime_dim, sdate_dim)) |
all(names(dim(Dates)) == c(sdate_dim, ftime_dim))) {
if (is.null(startdates)) {
startdates <- Subset(Dates, along = ftime_dim, 1, drop = 'selected')
} else if ((!inherits(startdates, "POSIXct") & !inherits(startdates, "Date")) &&
(!is.character(startdates) | (all(nchar(startdates) != 10) &
all(nchar(startdates) != 8) & all(nchar(startdates) != 6)))) {
warning("Parameter 'startdates' should be a character string containing ",
"the start dates in the format 'yyyy-mm-dd', 'yyyymmdd', 'yyyymm', ",
"'POSIXct' or 'Dates' class.")
startdates <- Subset(Dates, along = ftime_dim, 1, drop = 'selected')
}
} else {
stop("Parameter 'Dates' must have start date dimension and ",
"forecast time dimension.")
# startdates
if (is.null(startdates)) {
if (is.null(sdate_dim)) {
startdates <- 'XXX'
} else {
startdates <- rep('XXX', dim(data)[sdate_dim])
}
} else {
if (is.null(sdate_dim)) {
if (length(startdates) != 1) {
warning("Parameter 'startdates' has length more than 1. Only first ",
"value will be used.")
startdates <- startdates[[1]]
}
if (is.null(Datasets)) {
if (!single_file) {
warning("Parameter 'Datasets' is NULL. Files will be saved with a ",
"directory name of 'XXX'.")
}
Datasets <- rep('XXX', n_datasets )
}
if (inherits(Datasets, 'list')) {
Datasets <- names(Datasets)
if (n_datasets > length(Datasets)) {
warning("Dimension 'Datasets' in 'data' is greater than those listed in ",
"element 'Datasets' and the first element will be reused.")
Datasets <- c(Datasets, rep(Datasets[1], n_datasets - length(Datasets)))
} else if (n_datasets < length(Datasets)) {
warning("Dimension 'Datasets' in 'data' is smaller than those listed in ",
"element 'Datasets' and only the firsts elements will be used.")
Datasets <- Datasets[1:n_datasets]
# extra_string
if (!is.null(extra_string)) {
if (!is.character(extra_string)) {
stop("Parameter 'extra_string' must be a character string.")
}
}
alldims <- c(dat_dim, var_dim, sdate_dim, lon_dim, lat_dim, memb_dim, ftime_dim)
if (!all(dimnames %in% alldims)) {
unknown_dims <- dimnames[which(!dimnames %in% alldims)]
warning("Detected unknown dimension: ", paste(unknown_dims, collapse = ', '))
memb_dim <- c(memb_dim, unknown_dims)
alldims <- c(dat_dim, var_dim, sdate_dim, lon_dim, lat_dim, memb_dim, ftime_dim)
}
# Reorder
if (any(dimnames != alldims)) {
data <- Reorder(data, alldims)
## NetCDF dimensions definition
defined_dims <- NULL
if (is.null(Dates)) {
filedims <- dimnames[which(!dimnames %in% c(dat_dim, var_dim))]
} else {
filedims <- dimnames[which(!dimnames %in% c(dat_dim, var_dim, sdate_dim, ftime_dim))]
}
for (i_coord in filedims) {
dim_info <- list()
# vals
if (i_coord %in% names(coords)) {
dim_info[['vals']] <- as.vector(coords[[i_coord]])
} else {
dim_info[['vals']] <- 1:dim(data)[i_coord]
}
# name
dim_info[['name']] <- i_coord
# len
dim_info[['len']] <- as.numeric(dim(data)[i_coord])
# unlim
dim_info[['unlim']] <- FALSE
# create_dimvar
dim_info[['create_dimvar']] <- TRUE
## metadata
if (i_coord %in% names(metadata)) {
if ('variables' %in% names(attributes(metadata[[i_coord]]))) {
# from Start: 'lon' or 'lat'
attrs <- attributes(metadata[[i_coord]])[['variables']][[i_coord]]
i_coord_info <- attrs[!sapply(attrs, inherits, 'list')]
} else if (inherits(metadata[[i_coord]], 'list')) {
i_coord_info <- metadata[[i_coord]]
} else if (!is.null(attributes(metadata[[i_coord]]))) {
i_coord_info <- attributes(metadata[[i_coord]])
} else {
stop("Metadata is not correct.")
}
# len
if ('size' %in% names(i_coord_info)) {
if (i_coord_info[['size']] != dim(data)[i_coord]) {
i_coord_info[['original_len']] <- i_coord_info[['size']]
}
# units
if (!('units' %in% names(i_coord_info))) {
dim_info[['units']] <- ''
} else {
dim_info[['units']] <- i_coord_info[['units']]
}
# calendar
if (!('calendar' %in% names(i_coord_info))) {
dim_info[['calendar']] <- NA
} else {
dim_info[['calendar']] <- i_coord_info[['calendar']]
}
# longname
if ('long_name' %in% names(i_coord_info)) {
dim_info[['longname']] <- i_coord_info[['long_name']]
} else if ('longname' %in% names(i_coord_info)) {
dim_info[['longname']] <- i_coord_info[['longname']]
# or use_load
if (i_coord %in% .KnownLonNames()) {
dim_info[['longname']] <- 'longitude'
} else if (i_coord %in% .KnownLatNames()) {
dim_info[['longname']] <- 'latitude'
}
# extra information
if (!is.null(names(i_coord_info))) {
extra_info_dim[[i_coord]] <- i_coord_info
}
# units
dim_info[['units']] <- "adim"
# longname
new_dim <- list(ncdim_def(name = dim_info[['name']], units = dim_info[['units']],
vals = dim_info[['vals']],
unlim = dim_info[['unlim']],
create_dimvar = dim_info[['create_dimvar']],
# calendar = dim_info[['calendar']],
longname = dim_info[['longname']])
)
names(new_dim) <- i_coord
defined_dims <- c(defined_dims, new_dim)
defined_vars <- list()
if (!single_file) {
for (i in 1:n_datasets) {
path <- file.path(destination, Datasets[i], varname)
for (j in 1:n_vars) {
dir.create(path[j], recursive = TRUE)
startdates <- gsub("-", "", startdates)
dim(startdates) <- c(length(startdates))
if (is.null(dat_dim) & is.null(var_dim)) {
data_subset <- data
} else if (is.null(dat_dim)) {
data_subset <- Subset(data, c(var_dim),
list(j), drop = 'selected')
print(dim(data))
print(dat_dim)
# ERROR HERE in Subset
# data_subset <- Subset(data, along = c(dat_dim), list(i), drop = 'selected')
dim(data) <- dim(data)[-which(dimnames == dat_dim)]
data_subset <- data
print(dim(data_subset))
data_subset <- Subset(data, c(dat_dim, var_dim),
list(i, j), drop = 'selected')
}
if (is.null(Dates)) {
input_data <- list(data_subset, startdates)
target_dims <- list(c(ftime_dim, memb_dim, lat_dim, lon_dim),
} else {
input_data <- list(data_subset, startdates, Dates)
target_dims = list(c(ftime_dim, memb_dim, lat_dim, lon_dim),
Apply(data = input_data,
target_dims = target_dims,
fun = .saveExp,
destination = path[j],
defined_dims = defined_dims,
ftime_dim = ftime_dim,
varname = varname[j],
metadata_var = metadata[[varname[j]]],
}
}
} else {
# Datasets definition
if (!is.null(dat_dim)) {
new_dim <- list(ncdim_def(name = dat_dim, units = "adim",
vals = 1 : dim(data)[dat_dim],
longname = 'Datasets', create_dimvar = TRUE))
names(new_dim) <- dat_dim
defined_dims <- c(new_dim, defined_dims)
extra_info_dim[[dat_dim]] <- list(Datasets = paste(Datasets, collapse = ', '))
}
if (!is.null(Dates)) {
# sdate definition
sdates <- Subset(Dates, along = ftime_dim, 1, drop = 'selected')
differ <- as.numeric((sdates - sdates[1])/3600)
new_dim <- list(ncdim_def(name = sdate_dim, units = paste('hours since', sdates[1]),
vals = differ,
longname = sdate_dim, create_dimvar = TRUE))
names(new_dim) <- sdate_dim
defined_dims <- c(defined_dims, new_dim)
extra_info_dim[[sdate_dim]] <- list(sdates = paste(sdates, collapse = ', '))
# ftime definition
ftime_dates <- Subset(Dates, along = sdate_dim, 1, drop = 'selected')
differ <- as.numeric((ftime_dates - ftime_dates[1])/3600)
# DOUBT HERE: which values we take? # FIX Dates[1] for single_file = TRUE
dim_time <- list(ncdim_def(name = ftime_dim, units = paste('hours since', Dates[1]),
vals = differ, calendar = 'proleptic_gregorian',
longname = ftime_dim, unlim = TRUE))
names(dim_time) <- ftime_dim
defined_dims <- c(defined_dims, dim_time)
}
# var definition
defined_vars <- list()
for (j in 1:n_vars) {
i_var_info <- metadata[[varname[j]]][!sapply(metadata[[varname[j]]], inherits, 'list')]
# units
if ('units' %in% names(i_var_info)) {
var_info[['units']] <- i_var_info[['units']]
i_var_info[['units']] <- NULL
# dim
var_info[['dim']] <- defined_dims
# missval
if ('missval' %in% names(i_var_info)) {
var_info[['missval']] <- i_var_info[['missval']]
i_var_info[['missval']] <- NULL
} else {
var_info[['missval']] <- NULL
}
# longname
if (any(c('longname', 'long_name') %in% names(i_var_info))) {
longname <- names(i_var_info)[which(names(i_var_info) %in% c('longname', 'long_name'))]
var_info[['longname']] <- i_var_info[[longname]]
i_var_info[[longname]] <- NULL
} else {
}
# prec
if ('prec' %in% names(i_var_info)) {
var_info[['prec']] <- i_var_info[['prec']]
i_var_info[['prec']] <- NULL
} else {
prec <- typeof(data)
if (prec == 'character') {
var_info[['prec']] <- 'char'
}
if (any(prec %in% c('short', 'float', 'double', 'integer', 'char', 'byte'))) {
var_info[['prec']] <- prec
} else {
var_info[['prec']] <- 'double'
}
}
# extra information
if (!is.null(names(i_var_info))) {
extra_info_var[[varname[j]]] <- i_var_info
}
new_var <- list(ncvar_def(name = var_info[['name']],
units = var_info[['units']],
dim = var_info[['dim']],
missval = var_info[['missval']],
longname = var_info[['longname']],
prec = var_info[['prec']]))
defined_vars <- c(defined_vars, new_var)
}
if (is.null(extra_string)) {
file_name <- paste0(varname[j], "_", extra_string, ".nc")
}
full_filename <- file.path(destination, file_name)
file_nc <- nc_create(full_filename, defined_vars)
if (is.null(var_dim)) {
ncvar_put(file_nc, varname, vals = data)
for (j in 1:n_vars) {
ncvar_put(file_nc, defined_vars[[j]]$name,
vals = Subset(data, var_dim, j, drop = 'selected'))
# Additional dimension attributes
for (dim in names(defined_dims)) {
if (dim %in% names(extra_info_dim)) {
for (info_dim in names(extra_info_dim[[dim]])) {
ncatt_put(file_nc, dim, info_dim, as.character(extra_info_dim[[dim]][[info_dim]]))
}
}
}
# Additional dimension attributes
for (var in names(defined_vars)) {
if (var %in% names(extra_info_var)) {
for (info_var in names(extra_info_var[[var]])) {
ncatt_put(file_nc, var, info_var, as.character(extra_info_var[[var]][[info_var]]))
}
}
}
nc_close(file_nc)
.saveExp <- function(data, startdates = NULL, dates = NULL, destination = "./",
defined_dims, ftime_dim = 'time', varname = 'var',
metadata_var = NULL, extra_info_dim = NULL,
if (!is.null(dates)) {
differ <- as.numeric((dates - dates[1])/3600)
dim_time <- list(ncdim_def(name = ftime_dim, units = paste('hours since', dates[1]),
vals = differ, calendar = 'proleptic_gregorian',
longname = ftime_dim, unlim = TRUE))
names(dim_time) <- ftime_dim
defined_dims <- c(defined_dims, dim_time)
}
## Define var metadata
var_info <- NULL
extra_info_var <- NULL
i_var_info <- metadata_var[!sapply(metadata_var, inherits, 'list')]
# name
# units
if ('units' %in% names(i_var_info)) {
var_info[['units']] <- i_var_info[['units']]
i_var_info[['units']] <- NULL
} else {
var_info[['units']] <- ''
}
# dim
var_info[['dim']] <- defined_dims
# missval
if ('missval' %in% names(i_var_info)) {
var_info[['missval']] <- i_var_info[['missval']]
i_var_info[['missval']] <- NULL
} else {
var_info[['missval']] <- NULL
}
# longname
if (any(c('longname', 'long_name') %in% names(i_var_info))) {
longname <- names(i_var_info)[which(names(i_var_info) %in% c('longname', 'long_name'))]
var_info[['longname']] <- i_var_info[[longname]]
i_var_info[[longname]] <- NULL
} else {
if ('prec' %in% names(i_var_info)) {
var_info[['prec']] <- i_var_info[['prec']]
i_var_info[['prec']] <- NULL
prec <- typeof(data)
if (prec == 'character') {
var_info[['prec']] <- 'char'
}
if (any(prec %in% c('short', 'float', 'double', 'integer', 'char', 'byte'))) {
var_info[['prec']] <- prec
} else {
var_info[['prec']] <- 'double'
}
}
# extra information
if (!is.null(names(i_var_info))) {
extra_info_var <- i_var_info
datanc <- ncvar_def(name = var_info[['name']],
units = var_info[['units']],
dim = var_info[['dim']],
missval = -99999,
# longname = var_info[['longname']],
# prec = var_info[['prec']]
)
file_name <- paste0(varname, "_", startdates, ".nc")
file_name <- paste0(varname, "_", extra_string, "_", startdates, ".nc")
full_filename <- file.path(destination, file_name)
ncvar_put(file_nc, datanc, data)
for (dim in names(defined_dims)) {
if (dim %in% names(extra_info_dim)) {
for (info_dim in names(extra_info_dim[[dim]])) {
ncatt_put(file_nc, dim, info_dim, as.character(extra_info_dim[[dim]][[info_dim]]))
}
}
}
# Additional dimension attributes
if (!is.null(extra_info_var)) {
for (info_var in names(extra_info_var)) {
ncatt_put(file_nc, datanc, info_var, as.character(extra_info_var[[info_var]]))
if (use_load) {
ncatt_put(file_nc, 'longitude', 'gridtype', 'gaussian')
ncatt_put(file_nc, 'longitude', 'cdo_grid_name', 't426grid')
ncatt_put(file_nc, 'longitude', 'first_lon', 't426grid')
ncatt_put(file_nc, 'longitude', 'last_lon', 't426grid')
ncatt_put(file_nc, 'latitude', 'gridtype', 'gaussian')
ncatt_put(file_nc, 'latitude', 'cdo_grid_name', 't426grid')
ncatt_put(file_nc, 'longitude', 'first_lat', 't426grid')
ncatt_put(file_nc, 'longitude', 'last_lat', 't426grid')
}