write_ncdf: Export a data cube as netCDF file(s)

Description

This function will read chunks of a data cube and write them to a single (the default) or multitple (if chunked = TRUE) netCDF file(s). The resulting file(s) uses the enhanced netCDF-4 format, supporting chunking and compression.

Usage

write_ncdf(
  x,
  fname = tempfile(pattern = "gdalcubes", fileext = ".nc"),
  overwrite = FALSE,
  write_json_descr = FALSE,
  with_VRT = FALSE,
  pack = NULL,
  chunked = FALSE
)

Value

returns (invisibly) the path of the created netCDF file(s)

Arguments

x: a data cube proxy object (class cube)
fname: output file name
overwrite: logical; overwrite output file if it already exists
write_json_descr: logical; write a JSON description of x as additional file
with_VRT: logical; write additional VRT datasets (one per time slice)
pack: reduce output file size by packing values (see Details), defaults to no packing
chunked: logical; if TRUE, write one netCDF file per chunk; defaults to FALSE

Details

The resulting netCDF file(s) contain three dimensions (t, y, x) and bands as variables.

If write_json_descr is TRUE, the function will write an addition file with the same name as the NetCDF file but ".json" suffix. This file includes a serialized description of the input data cube, including all chained data cube operations.

To reduce the size of created files, values can be packed by applying a scale factor and an offset value and using a smaller integer data type for storage (only supported if chunked = TRUE). The pack argument can be either NULL (the default), or a list with elements type, scale, offset, and nodata. type can be any of "uint8", "uint16" , "uint32", "int16", or "int32". scale, offset, and nodata must be numeric vectors with length one or length equal to the number of data cube bands (to use different values for different bands). The helper function pack_minmax can be used to derive offset and scale values with maximum precision from minimum and maximum data values on original scale.

If chunked = TRUE, names of the produced files will start with name (with removed extension), followed by an underscore and the internal integer chunk number.

Examples

Run this code

# create image collection from example Landsat data only 
# if not already done in other examples
if (!file.exists(file.path(tempdir(), "L8.db"))) {
  L8_files <- list.files(system.file("L8NY18", package = "gdalcubes"),
                         ".TIF", recursive = TRUE, full.names = TRUE)
  create_image_collection(L8_files, "L8_L1TP", file.path(tempdir(), "L8.db"), quiet = TRUE) 
}

L8.col = image_collection(file.path(tempdir(), "L8.db"))
v = cube_view(extent=list(left=388941.2, right=766552.4, 
              bottom=4345299, top=4744931, t0="2018-04", t1="2018-04"),
              srs="EPSG:32618", nx = 497, ny=526, dt="P1M")
write_ncdf(select_bands(raster_cube(L8.col, v), c("B04", "B05")), fname=tempfile(fileext = ".nc"))