fnc_write: Function to reduce data output from automated LWFB90-Runs

In rhabel/modLWFB90: R-package for applying LWF-Brook90 directly from FVA-BW servers

Function to reduce data output from automated LWFB90-Runs

Description

LWFBrook90 creates a lot of output files. In order to keep data storage to a minimum, both run_LWFB90 and run_multisite_LWFB90 provide an output_fun - argument that can be used to reduce the output and directly write it to a database. This is what this function is made for.

LWFBrook90 creates a lot of output files. In order to keep data storage to a minimum, both run_LWFB90 and run_multisite_LWFB90 provide an output_fun - argument that can be used to reduce the output and directly write it. This is what this function is made for. Since v.0.5.0 LWFBrook90R changed the output/default settings.
fnc_write reduces and aggregates all columns returned by run_LWFB90 (see help page).
fnc_write_agg enables an even more detailed selection of drought indices. See detail section for a full list. Because many indices are calculated, this function will be significantly slower, so check, if you can get the necessary information from the standard list of output variables.
Both functions write .rds files with the desired output for each point. fnc_write_to_sql can be used to convert these files into a SQLite-DB.
This "step-in-between" is necessary because SQLite does not support parallel writing. Processed can be all output-values listed in run_LWFB90.

Usage

fnc_write(
  x,
  aggr_tp,
  out_tables,
  col_select_outputs = NA,
  col_select_layer = NA,
  depths = NA,
  dir_name = NA,
  soil_nm
)

fnc_write(
  x,
  aggr_tp,
  out_tables,
  col_select_outputs = NA,
  col_select_layer = NA,
  depths = NA,
  dir_name = NA,
  soil_nm
)

Arguments

x	one of the intermediate products of run_LWFB90 or run_multisite_LWFB90, which is further processed internally. Can't be adjusted.
aggr_tp	a string containing the desired aggregation time period. Can be daily, monthly, vegper, yearly, complete (average over complete time period modelled), or any combination of the four (i.e. monthly_vegper). The latter creates one result-table for each aggregation term detectable within the string. If aggregation periods other than daily are selected, all unitless balance terms of the "Outputs"-table of run_LWFB90 will be averaged using mean (i.e. relawat will be aggreated to "mean relawat within each years vegetation period"). All other terms will be summed up over the desired time period (i.e. rfal will be aggreated to "sum of rainfall within each years time period"). If aggregation periods other than daily are selected, all wetnes-related terms of the "Layer"-table of run_LWFB90 (swati, theta, wetnes and psimi) will be averaged using mean (i.e. relawat will be aggreated to "mean relawat within each years vegetation period in the respective layer"). All flow-related terms will be summed up over the desired time period (i.e. vrfl will be aggreated to "sum of vertical flow within each years time period").
out_tables	which tables of the run_LWFB90- tables will information be required from: layer, outputs, or outputs_layer
col_select_outputs	a string containing the desired columns from the "outputs"-outputs-list of run_LWFB90. BETTER MAKE A SELECTION or RAM and result-size will go through the roof! Plus be honest - you won't need everything.
col_select_layer	a string containing the desired columns from the "layer"-outputs-list of run_LWFB90. BETTER MAKE A SELECTION or RAM and result-size will go through the roof! Plus be honest - you won't need everything
depths	dephts for which layer-discrete results shall be processed. Must be a vector of depths in cm (i.e. c(10, 30, 60) ). If kept at NA, all depths will be processed, so better make a selection. If the respective soil data frame isn't deep enough to get all data requested here, all valid depths will be created.
dir_name	directory for tmp files, if NA as in default, results are returned to console
soil_nm	name of the modelled site. If the function is used in run_multisite_LWFB90 will be processed internally. Must be provided if only one site is modelled with run_LWFB90. This function argument will set the name of the result-files.
layercols	a sting containing the desired output products. Full list of possible output columns can be find on the help page of run_LWFB90 under Layer outputs
dailycols	same as layercols but for daily output. For possible options see run_LWFB90
db_name	name and file path of the SQL-database

Value

Returns the desired output to the database directly.

Returns the desired output to .rds files, or, the console

Output column selection

IT IS HIGHLY RECOMMENDED TO MAKE A SUBSELECTION, OR THERE WILL BE A LOT OF COLUMNS. For a complete list of possible output types plus description, see the details sections of run_LWFB90

Examples

# create simple sample data:
param_b90 <- set_paramLWFB90()
options_b90 <- set_optionsLWFB90()

# Set start and end Dates for the simulation
options_b90$startdate <- as.Date("2003-01-01")
options_b90$enddate <- as.Date("2006-12-31")

# Derive soil hydraulic properties from soil physical properties
# using pedotransfer functions
soil <- cbind(slb1_soil, hydpar_wessolek_tab(slb1_soil$texture))

# Run LWF-Brook90
x <- run_LWFB90(

  # run_LWFB90-function arguments:
  options_b90 = options_b90,
  param_b90 = param_b90,
  climate = slb1_meteo,
  soil = soil,

  # keep it easy to read
  rtrn_input = F,
  rtrn_output = F,

  output_fun = fnc_write,
  # fnc_write-function arguments:

  aggr_tp = c("daily_monthly_vegper_yearly_complete"),                            # all aggregation options for demonstration purposes
  out_tables = c("outputs_layer"),                                                # all output sources for demonstration purposes
  col_select_outputs =  c("yr", "mo", "da", "doy", "stres", "relawat", "vrfln"),  # subset from both mean and sum-aggregation
  col_select_layer = c("yr", "mo", "da", "doy", "nl", "theta", "swati", "tran"),  # subset from both mean and sum-aggregation
  depths = c(10,30,600),                                                          # last one won't be calculated
  dir_name = NA,                                                                  # return to console
  soil_nm = "test_soil"
)



# multiple Plots:
# set soil-list
ls.soil <- fnc_get_soil(df.ids = test.ids.bds,
                        soil_option = "BZE",
                        PTF_to_use = "WESSOLEK",
                        add_BodenInfo = F)
ls.parms <- fnc_get_params(df.ids = test.ids.bds,
                           tree_species = "spruce")
ls.clim <- fnc_get_clim(df.ids = test.ids.bds,
                        mindate = as.Date("2002-01-01"),
                        maxdate = as.Date("2004-12-31"))

# run multiple points back to console:
results <- run_multisite_LWFB90(options_b90 = options_b90,
                            param_b90 = ls.parms,
                            soil = ls.soil,

                            climate = ls.clim,


                            all_combinations = F,
                            cores = 5,

                            rtrn_output = F,
                            rtrn_input = F,


                            output_fun = fnc_write,

                            aggr_tp = c("daily_monthly_vegper_yearly_complete"),                            # all aggregation options for demonstration purposes
                            out_tables = c("outputs_layer"),                                                # all output sources for demonstration purposes
                            col_select_outputs =  c("yr", "mo", "da", "doy", "stres", "relawat", "vrfln"),  # subset from both mean and sum-aggregation
                            col_select_layer = c("yr", "mo", "da", "doy", "nl", "theta", "swati", "tran"),  # subset from both mean and sum-aggregation
                            depths = c(10,30,600),                                                          # last one won't be calculated

                            dir_name= NA                                                            # write to ./tmp/

                            )
results[[1]]

# run multiple points and write output
run_multisite_LWFB90(options_b90 = options_b90,
                            param_b90 = ls.parms,
                            soil = ls.soil,

                            climate = ls.clim,


                            all_combinations = F,
                            cores = 5,

                            rtrn_output = F,
                            rtrn_input = F,


                            output_fun = fnc_write,

                            aggr_tp = c("daily_monthly_vegper_yearly_complete"),                            # all aggregation options for demonstration purposes
                            out_tables = c("outputs_layer"),                                                # all output sources for demonstration purposes
                            col_select_outputs =  c("yr", "mo", "da", "doy", "stres", "relawat", "vrfln"),  # subset from both mean and sum-aggregation
                            col_select_layer = c("yr", "mo", "da", "doy", "nl", "theta", "swati", "tran"),  # subset from both mean and sum-aggregation
                            depths = c(10,30,600),                                                          # last one won't be calculated

                            dir_name= "./tmp/"                                                              # write to ./tmp/

                            )


# delete example tmp file
unlink("./tmp/", recursive = T)

rhabel/modLWFB90 documentation built on Jan. 21, 2025, 8:04 p.m.