makeAggregatedDataset: Create a dataset from a collection of (netCDF) files

In SantanderMetGroup/loadeR: A climate4R package for data access <http://meteo.unican.es/climate4r>

Create a dataset from a collection of (netCDF) files

Description

Creates virtual datasets by modifying and combining other datasets via NcML.

Usage

makeAggregatedDataset(
  source.dir,
  ncml.file,
  file.ext = "nc",
  aggr.dim = "time",
  pattern = NULL,
  recursive = FALSE,
  verbose = TRUE,
  timeUnitsChange = TRUE
)

Arguments

source.dir	Parent directory containing the files to be aggregated
ncml.file	Full path of the output NcML file
file.ext	Character string indicating the extension of the CDM datasets to be aggregated. Default to nc (netCDF).
aggr.dim	Character string indicating the dimension along which the files will be concatenated. Default to "time".
pattern	An optional regular expression. Only file names which match the regular expression will be considered in the aggregation (see regexp). This argument can be useful in order to save time when only a particular subset of variables from the whole collection is needed. Default to NULL, meaning that all files in the search path are included (See next argument). See Notes.
recursive	Logical. Should the listing of files to be aggregated recurse into directories?. Default to FALSE. This is useful for instance when each variable is stored in a sepparate subdirectory.
verbose	Logical. Should additional information of the NcML file creation steps be printed on screen?. Default to TRUE.
timeUnitsChange	Logical. Set the timeUnitsChange option, useful when aggregating many files across the time dimension which have relative time axis. Default to TRUE.

Details

The NetCDF Markup Language (NcML) is an XML dialect that allows creating CDM datasets (i.e.: any collection of scientific data which can be accessed through the NetCDF-Java / CDM library). The NcML document refers to another dataset called the referenced CDM dataset, generally composed on a number of netCDF files (but also grib, hdf or many other binary file formats) containing the geo-referenced data. This function creates a NcML file from multiple CDM files that are conveniently combined ("aggregated") along the selected dimension. The use of NcML is not only intended for CDM file combination, and this function considers just a particular case. NcML is a powerful, yet relatively simple way of dealing with large, complex datasets in a straightforward manner. Among other capabilitities, NcML files can be generated in order to add/delete metadata and variables to be renamed, added, deleted and restructured. This function is intended for a simple operation of aggregation of collections of netCDF files, as it is the most common case of gridded climate datasets.

Value

Creates a NcML file at the specified location.

Note

The current implementation of the function only considers datasets in which each file stores one single variable. For other dataset configurations, please refer to the NcML tutorial.

Wildcards (or globbing) patterns as used by most shells can be mapped to regular expression in R using glob2rx.

Author(s)

J. Bedia, with contributions of E. Tourigny and A. Cofiño

References

NcML Tutorial http://www.unidata.ucar.edu/software/thredds/current/netcdf-java/ncml/Tutorial.html. (Last accessed 14 Mar 2016.).

See Also

Other loading: loadCircularGridData(), loadGridData(), loadStationData()

SantanderMetGroup/loadeR documentation built on July 4, 2023, 4:29 a.m.