Skip to contents

Encode and decode full parameter names

Source: R/param_names.R
param_names.Rd

A calibratable parameter is identified by its group, name and index (the columns of parameters()). These helpers turn that triple into a single string – the full name, group/name[index] – and back. Calibration and sensitivity tools use the full name as a key, for example in PEST++ control files and results tables, where a parameter has to be one token.

The format is fixed because the strings are stored: a missing group or index is written as NA, so NA/inflow/1 encodes to "NA/inflow[1]" and NA/ce/NA to "NA/ce[NA]".

Decoding splits at the first / and the last [...]. The group therefore cannot itself contain a /, and a group literally called "NA" decodes as a missing group. A name may contain / (for example "mixing/coef_mix_conv") and [.

Usage

encode_param(group, name, index)

decode_param_full(name_full)

decode_param(name_full)

Arguments

group

character; the parameter group (NA for none).

name

character; the parameter name.

index

integer; the index of the parameter within a vector (NA for none).

name_full

character; full parameter names in the format group/name[index], as made by encode_param().

Value

encode_param() returns a character vector of full names, recycled like paste0(); character(0) when group, name and index are all empty.

decode_param_full() returns a data.frame with columns group, name and index, one row per element of name_full.

decode_param() returns the name part only. Elements that are not in the group/name[index] format (and NA) decode to NA.

Examples

nm <- encode_param(group = c("light", NA), name = c("Kw", "inflow"),
                   index = c(1L, NA))
nm
#> [1] "light/Kw[1]"   "NA/inflow[NA]"
decode_param(nm)
#> [1] "Kw"     "inflow"
decode_param_full(nm)
#>   group   name index
#> 1 light     Kw     1
#> 2  <NA> inflow    NA