Skip to contents

Generate Catch-Level Validation Flags

get_catch_flags.Rd

This function validates catch-level data from KEFS surveys and generates alert flags for inconsistencies in sample weights relative to total catch weights. It focuses on submissions where catch outcome was "yes" and checks for logical inconsistencies in the relationship between sample weights and total weights.

Usage

get_catch_flags(dat = NULL, max_length_cm = 500)

Arguments

dat

A data frame containing KEFS survey catch data with columns:

  • submission_id: Unique identifier for the submission

  • catch_outcome: Whether catch was recorded ("yes" or "no")

  • total_sample_weight: Total weight of all samples in kilograms

  • total_catch_weight: Total weight of the entire catch in kilograms

  • sample_weight: Weight of individual sample in kilograms

  • sample_price: Price of the sample

  • length_max_cm: Largest fish measured for the row's species, if any

max_length_cm

Numeric. Absolute upper bound on length_max_cm, in centimetres. Defaults to 500. The largest fish is tested rather than the mean, which one impossible fish among normal ones can stay under.

Value

A data frame with columns:

  • submission_id: The original submission identifier

  • alert_flag_catch: Comma-separated string of alert codes (NA if no alerts):

    • "5.1": Total sample weight exceeds total catch weight

    • "5.2": Individual sample weight exceeds either total sample weight or total catch weight

    • "5.3": Length exceeds max_length_cm

  • alert_catches: For 5.2 and 5.3, the species sample each was raised on: its number in the form (catch_number, 1 for the first species), code, species and the value recorded. NULL when none was.

Details

The function only processes submissions where catch_outcome is "yes". It identifies two types of weight inconsistencies:

  1. The sum of all samples exceeds the total reported catch

  2. An individual sample weighs more than the total it's supposed to be part of

Alert 5.3 is an absolute ceiling, not a per-species check. Zanzibar and Mozambique bound length per taxon from FishBase morphometrics; Kenya has no morphometrics table, so there is no species-specific bound to apply here. The default only catches the physically impossible

Examples

if (FALSE) { # \dontrun{
catch_flags <- get_catch_flags(dat = survey_data)
} # }