Pre-process WorldFish Zanzibar Gleaning Surveys

Downloads raw structured gleaning survey data from cloud storage and preprocesses it into a single analysis-ready data frame. The function assembles three pieces and joins them on the submission:

1. General info – strips the Kobo group prefixes (group_general/, group_trip/, no_fishers/, demographics/, group_gleaning_activity/, group_supply_chain/), selects and renames the trip, demographic, activity and supply-chain fields, coalesces the conditional landing_site columns into one, and coerces dates and numeric fields.

2. Catch info – reshapes the wide group_catch block into a tidy long table (one row per submission x shell group x size class) via reshape_gleaning_catch, unifying the parallel bucket/plastic container fields and applying conservative sanitisation.

3. Catch totals – per submission, sums individuals across size classes (total_individuals) and reconstructs catch weight as unit_weight_kg * n_containers (total_catch_kg); the container fields are constant within a submission, hence first().

Usage

preprocess_wf_gleaning(log_threshold = logger::DEBUG)

Arguments

log_threshold

Logging threshold level (default: logger::DEBUG).

Value

A data frame of preprocessed gleaning surveys: one row per submission x shell group x size class, with general/demographic/activity/ supply-chain fields plus total_individuals and total_catch_kg.

Details

Configurations are read from config.yml with the following necessary parameters:

surveys:
  wf_gleaning:
    raw:
      file_prefix:
      version:
storage:
  google:
    key:
    options:
      project:
      bucket:
      service_account_key:

The function uses logging to track progress.

See also

reshape_gleaning_catch, sanitize_gleaning_inputs