Editing Selection files

RS Minerve can plot simulation results and print result files for selections of simulation results. These selections can be edited in the user interface, saved to a chk file and also loaded from chk files. As the manual editing of the plot selection can be tedious for large models, it can be useful to automate the reading and writing of the chk files in R. The present vignette demonstrates how to do this.

library(riversCentralAsia)

Reading a chk file

filepath <- if (file.exists("../tests/testthat/test_selection.chk")) {
   "../tests/testthat/test_selection.chk"
} else {
  "https://raw.githubusercontent.com/hydrosolutions/RSMinerveR/main/tests/testthat/test_selection.chk"
}

xmldata <- readSelectionCHK(filepath)

The name of the selection:

selection_name <- xmldata[[1]]
print(selection_name)
#> [1] "New selection"

The description of the objects in the selection:

selection_data <- xmldata[[2]]
print(selection_data)
#> # A tibble: 29 × 4
#>    Model       Object     ID                   Variable                
#>    <chr>       <chr>      <chr>                <chr>                   
#>  1 Model Koksu Source     QSpring              Kichkinesay - QUp (m3/s)
#>  2 Model Koksu Comparator Comparator 1         QReference (m3/s)       
#>  3 Model Koksu Comparator Comparator 1         QSimulation (m3/s)      
#>  4 Model Koksu HBV92      Koksu downstream_eb1 Qr (m3/s)               
#>  5 Model Koksu HBV92      Koksu downstream_eb1 Qu (m3/s)               
#>  6 Model Koksu HBV92      Koksu downstream_eb1 Ql (m3/s)               
#>  7 Model Koksu HBV92      Koksu downstream_eb1 ETR (m/s)               
#>  8 Model Koksu HBV92      Koksu downstream_eb1 Peq (m/s)               
#>  9 Model Koksu HBV92      Koksu downstream_eb1 SWE (m)                 
#> 10 Model Koksu HBV92      Koksu downstream_eb1 WH (-)                  
#> # … with 19 more rows

Write a new selection

New selection name

new_selection_name <- "SOCONT Qtot"

Edit the selection to include all data for variable Qtot of all SOCONT objects in the model. To get a list of all SOCONT objects the user can read a parameter file or a file of initial conditions of the model. Another possibility is to read the gis layer with the hydrological response units used for the semi-automatic model creation in RSMinerve.
Let’s read the names of the SOCONT objects from the parameter file of one of the tutorials. The procedure is, of course, analogous for any other RSMinerve Object for which data should be extracted.

parfilepath <- if (file.exists("../tests/testthat/Tutorial_Parameters.txt")) {
   "../tests/testthat/Tutorial_Parameters.txt"
} else {
  "https://raw.githubusercontent.com/hydrosolutions/RSMinerveR/main/tests/testthat/Tutorial_Parameters.txt"
}
par <- readRSMParameters(parfilepath) |> 
  dplyr::filter(Object == "SOCONT")
Object_IDs <- unique(par$Name)
print(Object_IDs)
#> [1] "SOCONT 1" "SOCONT 2" "SOCONT 3"

Now construct a tibble of the variables for each object we’d like to select results from.

new_selection <- tibble::tibble(
  Model = rep("Tutorial_Model", length(Object_IDs)), 
  Object = rep("SOCONT", length(Object_IDs)), 
  ID = Object_IDs, 
  Variable = rep("Qtot (m3/s)")
)
print(new_selection)
#> # A tibble: 3 × 4
#>   Model          Object ID       Variable   
#>   <chr>          <chr>  <chr>    <chr>      
#> 1 Tutorial_Model SOCONT SOCONT 1 Qtot (m3/s)
#> 2 Tutorial_Model SOCONT SOCONT 2 Qtot (m3/s)
#> 3 Tutorial_Model SOCONT SOCONT 3 Qtot (m3/s)

Now write the data to a new file. In the codechuck below you will need to replace <filepath> with an actual path to a file to write.

writeSelectionCHK(<filepath>, new_selection, new_selection_name)