Compute a custom priority score beyond standard severity classification. For project-specific metrics like cost-effectiveness, species-weighted priority, or multi-criteria ranking.
Usage
lnk_score_custom(
conn,
crossings,
rules,
col_id = "modelled_crossing_id",
col_score = "priority_score",
to = NULL,
verbose = TRUE
)Arguments
- conn
A DBI::DBIConnection object.
- crossings
Character. Schema-qualified crossings table.
- rules
Named list of scoring rule specs. Each rule is a named list with:
- col
Column to evaluate (required).
- weight
Numeric weight (default 1).
- direction
"higher"(default) or"lower"is better.- sql
Optional raw SQL expression instead of
col. Developer API — must not contain user input.
- col_id
Character. Primary key column in the crossings table. Used for joining scores back to rows.
- col_score
Character. Name of output score column.
- to
Character. If
NULL, updates in-place. Otherwise writes to new table.- verbose
Logical. Report score distribution summary.
Details
Composable: severity from lnk_score_severity() is one input.
Upstream habitat value (from lnk_habitat_upstream()) is another.
Custom scoring combines them into a single priority number.
Weighted rank: each rule produces a rank (1 = best), multiplied by its weight, summed into a composite score. Lower composite = higher priority. This avoids unit-mixing problems (you can't add metres of habitat to severity categories, but you can add their ranks).
Examples
# --- "Which 10 crossings should we fix first?" ---
if (FALSE) { # \dontrun{
conn <- lnk_db_conn()
# Score severity first
lnk_score_severity(conn, "working.crossings")
# Then add upstream habitat (from fresh output)
lnk_habitat_upstream(conn, "working.crossings", "fresh.habitat")
# Now rank: severity weight 2x, habitat weight 3x
lnk_score_custom(conn, "working.crossings",
rules = list(
severity = list(col = "severity", weight = 2,
sql = "CASE severity
WHEN 'high' THEN 3
WHEN 'moderate' THEN 2
ELSE 1 END"),
habitat = list(col = "spawning_km", weight = 3)))
# Priority score distribution:
# min: 5.0 median: 12.3 max: 42.7
#
# Lower score = higher priority for remediation.
# Top 10: SELECT * FROM working.crossings
# ORDER BY priority_score LIMIT 10
} # }