Skip to contents

Validate a shinyplanr feature dictionary (Dict_Feature.csv)

Source: R/validate_config.R
validate_dict.Rd

Runs structural checks on the raw (unfiltered) feature dictionary read from Dict_Feature.csv in setup/3_setup_app.R, before the includeApp filter is applied. Call this immediately after readr::read_csv() and before dplyr::filter(includeApp).

Usage

validate_dict(Dict, strict = TRUE)

Arguments

Dict

A data frame. The raw (unfiltered) feature dictionary, typically the direct output of readr::read_csv(file.path(setup_dir, "Dict_Feature.csv")).

strict

Logical. If TRUE (default), stops immediately with a clear, actionable error message on the first failed check. If FALSE, all checks are run and a summary report is returned invisibly; warning() is called for each failure.

Value

When strict = FALSE, invisibly returns a named list of logical values (TRUE = passed, FALSE = failed) for each check. When strict = TRUE, returns invisible(TRUE) if all checks pass.

Details

Catching problems here – before the data is loaded – gives the deployer the clearest possible error messages, because the issue is in the CSV they just edited rather than buried inside a spatial data pipeline.

Checks performed

  • All required columns are present in Dict.

  • includeApp and includeJust columns are logical (TRUE/FALSE), not character or integer. A common mistake is editing the CSV in Excel, which can convert TRUE to 1 or "TRUE" (character), causing dplyr::filter(includeApp) to silently drop all rows.

  • All values in the type column are from the known set ("Feature", "Cost", "LockIn", "LockOut", "Bioregion", "EcosystemServices", "Justification"). A typo like "feature" (lowercase) silently excludes a row from all app processing.

  • nameVariable is unique within each type. Duplicates cause silent bugs in prioritizr (duplicate feature columns) and duplicate slider input IDs in the Shiny UI. Note: the same nameVariable may legitimately appear in both "LockIn" and "LockOut" rows (e.g. MPAs) – uniqueness is only enforced within each type.

  • At least one row has includeApp == TRUE and type == "Feature". An app with no active features cannot run a prioritisation.

  • All rows with includeApp == TRUE and type == "Feature" have targetMin, targetMax, and targetInitial values in the 0–100 range. Out-of-range values cause prioritizr to error at solve time.

Examples

if (FALSE) { # \dontrun{
# In setup/3_setup_app.R, immediately after reading the CSV:
Dict_raw <- readr::read_csv(file.path(setup_dir, "Dict_Feature.csv"))
shinyplanr::validate_dict(Dict_raw)

Dict <- Dict_raw |>
  dplyr::filter(includeApp) |>
  dplyr::arrange(type, categoryID, nameCommon)
} # }