- site_name and run_name are printed to console in
growR_run_loop
. - PhenologicalAutocut
- autocut algorithm is now outsourced to its own R6class.
- An additional option for determination of growing season start is available
through
ModvegeSite$set_SGS_method. This comes with a different available definition of the temperature sum and is reflected in the new methods
start_of_growing_season,
start_of_growing_season_mtd,
weighted_temperature_sum`.
j_start_of_growing_season
is written to output file headers.- Output file header contains
cut_DOYs
. growR_run_loop
allows use of previous year's state variables as initial conditions for current year by means of theindependent
argument.stubble_height
is now a parameter (i.e. a field ofModvegeParameters
) instead of a field ofMovegeSite
.- Lower biomass limit is only applied after
ST2
.
- Window sizes in multicriterial thermal definition were incorrect.
- Calculation of
fAgeDR
was incorrect. - Biomass is correctly updated after cuts.
dBM
can become negative, as it should.fW
now uses correct threshold value of 3.8 instead of 3.81.- Correction of AEv calculation.
- NULL value for simulations without cut_DOYs still produces a header in the output file.
- Bug where error message was not printed correctly in case of missing parameters.
- Biomass (BM, BMxx) was unaligned with biomass change (dBM, dBMxx).
- Green biomass has a lower bound defined by a minimum grass height. This is to prevent the population from completely dying out and effectively simulates reserves.
- The start of the growing season is now determined by the multicriterial thermal definition as described in chapter 2.3.1.3 of the dissertation of Andreas Schaumberger: Räumliche Modelle zur Vegetations- und Ertragsdynamik im Wirtschaftsgrünland, 2011, ISBN-13: 978-3-902559-67-8
- compare.R script is not fetched by
setup_directory()
.
-
Input integrity checking for WeatherData and ManagementData.
-
PScanPlotter (
plot_parameter_scan
) now allows to select which variable to display (from dBM, cBM and cBM_tot). -
Instances where simulation results for variable
dBM
are visualized or compared to experimental data now allow specifying and integration window fordBM
, which should be set to match the experimental reality. This concernsModvegeSite$plot()
andanalyze_parameter_scan()
as well as thecompare.R
script. -
S3 dispatch for plot method of
ModvegeSite
objects ->plot(mvs)
is now possible ifmvs
is aModvegeSite
instance. -
ParameterData
input checking: throws error on duplicate input parameter name. -
Debug utility conveniences
browse
andbrowse_end
. -
ModvegeSite$plot_XXX
functions for more insights into model behaviour, with XXX inwater
,limitations
,growth
andplot_var
for generic variable plotting.
-
Input data CSV files are now actual CSV files, instead of "semicolon-separated value" files.
-
ParameterData
now assumes default values for most parameters.check_parameters
now only throws an error if any of the really
required parameters are missing. -
Number of days per year is now inferred from weather data instead of hardcoded to 365, allowing simulations to be run for years with incomplete data.
-
Crop coefficient (
parameters$crop_coefficient
) and senescence capping are now model parameters instead of being hardcoded.
-
autocut:
get_annual_gross_yield
was incorrectly hardcoded to return 1. -
ParameterData$set_parameters
now updates initial condition values, if applicable.
-
Removed superfluous weather inputs.
-
Removed automatic temperature correction.
-
Redundant argument store_results in
growR_run_loop
. -
SEA
symmetrization around 1 is not enforced anymore.
- Initial CRAN submission.
- Renaming from
rmodvege
togrowR
.
- Port of
rmodvege-scripts
to R package structure.