getdata: Extracting Data from NetCDF File Connection
In retostauffer/Rcost: Cost 733 WLK Classification in R

Description Usage Arguments Details Value Author(s) Examples

Built to extract certain variables/forecast times from NetCDF files.

1 2	## Basic usage getdata(nc, init, varname, steps, level = NULL, subset = NULL, silent = FALSE )

`nc`	Open `ncdf4` connection
`init`	`Date`, `POSIXt`, or `character` object which can be converted into `POSIXt`. Date/time when the forecast was initialized. Crucial to compute the forecast steps (lead times).
`varname`	`character`, name of the variable to load.
`steps`	Optional. If not set all forecast steps from the `nc` file will be returned. Single or multiple steps can be defined (`numeric`).
`level`	Used when loading variables from pressure level files where the level has to be specified as a `numeric` value.
`subset`	Optional, default `NULL`. If not set the whole grid from the `nc` file will be returned. Can also be a `extent` object or a `RasterLayer` object. See details.
`silent`	`logical`, default is `FALSE`. If `TRUE` some messages will be suppressed.

subset allows to only load parts of the data from the nc connection. If subset is of class Extent (see extent) then only (a spatial) subset of the data will be taken. If subset is of class RasterLayer (see raster) the extent of the RasterLayer will be used to crop the data, the resolution (or grid specification) of this RasterLayer will be used to re-sample the data. This allows to change the resolution when loading the data.

Returns a RasterLayer (if only one step) is loaded or a RasterStack object where each layer corresponds to a forecast step. Names of the layers contain the initial initial date plus the forecast step.

Reto Stauffer

## Not run: 
library('maps'); library("raster")

# Open netcdf connections
init <- as.POSIXct("2011-11-18 00:00")
init_str <- strftime(init,"
ncs <- nc_open(sprintf("costgrib/ECMWF_cost_sfc_
ncp <- nc_open(sprintf("costgrib/ECMWF_cost_pl_

# Loading "tcw" from surface file, step 24, full field
x <- getdata(ncs,init,"tcw",24)
plot(x); map(add=T)

# Using an extent subset and loading three different steps
x <- getdata(ncs,init,"tcw",c(12,24,48),subset=extent(c(10,30,40,50)))
plot(x)

# Loading pressure level data (u on 700 hPa) and resample the data
# onto a coarser resolution.
template <- raster(xmn=9.5,xmx=30.5,ymn=39.5,ymx=50.5,nrows=21,ncols=11,
               crs=crs("+proj=longlat +datum=WGS84 +ellps=WGS84 +no_defs +towgs84=0,0,0"))
x <- getdata(ncp,init,"u",144,level=700,subset=template)
plot(x); map(add=T)

# Close connections
nc_close(ncs)
nc_close(ncp)

## End(Not run)