Skip to contents

Establish a DuckDB Connection to DATRAS Datasets

dr_con.Rd

This function creates a DuckDB connection to a specified DATRAS dataset type, facilitating access to trawl survey data stored in Parquet format. The dataset type determines which data is loaded from the remote source.

Usage

dr_con(
  type = NULL,
  trim = TRUE,
  url = "https://heima.hafro.is/~einarhj/datras",
  quiet = TRUE
)

Arguments

type

A character string specifying the dataset type. Available values (tables):

  • "HH": Haul-level data.

  • "HL": Catch-at-length data (filterable via the trim option).

  • "CA": Catch-at-age data (filterable via the trim option).

  • "species": Species dataset derived from ICES SpecWoRMS.

trim

Logical. For "HL" or "CA", if TRUE (default), non-essential fields are excluded. Ignored for other datasets.

url

URL to the Parquet file directory, currently defaulting to "https://heima.hafro.is/~einarhj/datras".

quiet

Logical. If TRUE (default), suppresses connection warnings and messages.

Value

A DuckDB dataset table.

Dataset Types

This function operates on the following dataset types:

  • "HH" (Haul-Level Data): Contains information related to individual haul events.

  • "HL" (Catch-at-Length Data): Records catches categorized by length class.

  • "CA" (Catch-at-Age Data): Includes age-based biological data (e.g., liver weight, length).

  • "species" (Species List): Derived from the ICES vocabulary 'SpecWoRMS' and includes species names and related metadata.

Dataset Paths

The dataset is accessed via HTTP/HTTPS paths at a user-defined or default URL location. The file names are inferred from the provided type parameter (e.g., a Parquet file named "HH.parquet" for "HH" type data).

Unique Identifier (.id)

For dataset types "HH", "HL", and "CA", a unique identifier column (.id) represent catenation of fields Survey, Year, Quarter, Country, Platform, Gear, StationName and HaulNumber seperated by ":" (see dr_add_id).

Examples

if (FALSE) { # \dontrun{
  # Establish connections
  dr_con("HH")                   # Connect to haul-level data.
  dr_con("HL", trim = FALSE)     # Include all fields for catch-at-length data.
  species_data <- dr_con("species")

  # Inspect species data
  dplyr::glimpse(species_data)
} # }