Generates an object of class eupp_config
Containing all the required
settings for downloadin data set inventories (index files) and the data
itself.
character of length 1
(see Details).
character of length 1
or NULL
(see Details).
character of length 1
(see Details).
an object which can be converted to POSIXct
. Character string
in ISO representation or an object of class Date
or POSIXt
.
NULL
(default) or character. When NULL
all available
parameters will be processed. Can be used to retrieve/process subsets.
NULL
(default) or integer. If set to NULL
all available forecast
steps will be processed. An integer sequence can be provided to only process
specific forecast steps; given in hours (e.g., c(6, 12)
for +6
and +12
hour ahead forecasts).
NULL
(default) or integer. If set to NULL
all available forecast
members will be processed. An integer sequence can be provided to only process
specific forecast members.
NULL
(default) or an object of class bbox
. Used for spatial subsetting.
NULL
or character of length 1
pointing to an existing
directory. Is used for data caching (caching grib index information/inventories)
which can be handy to save some time.
integer lengt 1
; version of the data set. Defaults to 0L
.
Typically not changed by an end-user.
Returns an object of class eupp_config
.
Input argument product
:
"reforecast"
: Reforecasts (or hindcasts).
"forecast"
: Forecast data.
"analysis"
: Analysis data; based on ECMWF ERA5 used as ground truth.
Input argument level
. Will be ignored if type = "efi"
(has no level).
"efi"
: Extreme forecast index (in this situation type
will be ignored).
"surf"
: Surface data.
"pressure"
: Pressure level data.
Input argument type
:
"ens"
: Ensemble members including control run (allows for optional argument members
).
"hr"
: High-resolution forecast.
Note that reforecasts always only become available on Mondays and Thursdays.
If type = "reforecast"
and the date
does not point to Mon/Thu
the function will throw an error.
TODO(R): This is not true, this happens when processing the data.
Ensemble forecasts (type = "ens"
consists of both, the ensemble control
run plus 50
ensemble members if members
is not specified.
Same is true for reforecasts except that it only consists of 20
ensemble
members. Possible combinations:
ID | product | level | type |
1 | analysis | surface | NULL |
2 | pressure | NULL | |
11 | forecast | efi | |
12 | surface | ens | |
13 | hr | ||
14 | pressure | ens | |
15 | hr | ||
21 | reforecast | surface | ens |
22 | pressure | ens |
# Analysis
c1 <- eupp_config("analysis", "surface", date = "2017-01-01", steps = 12)
c2 <- eupp_config("analysis", "pressure", date = "2017-01-01", steps = 12)
# Forecasts
c11 <- eupp_config("forecast", "efi", date = "2017-01-01", steps = 12)
c12 <- eupp_config("forecast", "surface", "ens", date = "2017-01-01", steps = 12)
c13 <- eupp_config("forecast", "surface", "hr", date = "2017-01-01", steps = 12)
c14 <- eupp_config("forecast", "pressure", "ens", date = "2017-01-01", steps = 12)
c15 <- eupp_config("forecast", "pressure", "hr", date = "2017-01-01", steps = 12)
# Reforecasts
c21 <- eupp_config("reforecast", "surface", "ens", date = "2017-01-02", steps = 12)
c22 <- eupp_config("reforecast", "pressure", "ens", date = "2017-01-02", steps = 12)