Skip to contents

Linear Model for Catch

Source: R/catch_lm.R
catch_lm.Rd

First stage regression model for catch.

Usage

catch_lm(
  dat,
  project,
  catch.formula,
  zoneID = NULL,
  exp.name = NULL,
  new.name = NULL,
  date,
  output = "matrix"
)

Arguments

dat

Primary data containing information on hauls or trips. Table in FishSET database contains the string 'MainDataTable'.

project

String, name of project.

catch.formula

A formula object specifying the linear model.See stats::lm().

zoneID

zone ID Variable in dat that identifies the individual zones or areas. Required if merging expected catch into dat using exp.name, or when creating a new expected catch matrix and exp.name is NULL (see output below).

exp.name

Name(s) of expected catch matrix to merge into dat.

new.name

Optional, string. When output = 'matrix', new.name will become the name of the new expected catch matrix saved to the FishSET DB expected catch list. When output = 'dataset', new.name will become the name of the new expected catch variable added to the primary dataset.

date

Date variable from dat used to create expected catch matrix.

output

Whether to output dat with the expected catch variable added ('dataset') or to save an expected catch matrix to the expected catch FishSET DB table ('matrix'). Defaults to output = 'matrix'.

Value

catch_lm() has two output options: dataset and matrix. When output == 'dataset', the primary dataset will be returned with the fitted values from the model added as a new column. The new column is named using new.name.

When output == 'matrix' an expected catch matrix is created and saved to the FishSET DB expected catch list (it is not outputted to the console). There are two ways to create an expected catch matrix: by using an existing expected catch matrix in catch.formula, or by using a zone-identifier column (i.e. zoneID) in the catch.formula. For example, if you have created an expected catch matrix named 'user1' using create_expectations(), catch.formula could equal catch ~ vessel_length * user1. In this case exp.name would equal 'user1'. Alternatively, you could create an expected catch matrix by specifying catch.formula as catch ~ vessel_length * zone. In this case, exp.name = NULL and zoneID = 'zone'.

Details

catch_lm() can merge an expected catch matrix into the primary dataset before running the linear model. This is done using by passing exp.name and zoneID to merge_expected_catch() and is for convenience; users can do this separately using merge_expected_catch() if desired, just make sure to leave exp.name empty before running catch_lm(). Merging expected catch in a separate step is useful for creating tables and plots before running a first stage linear regression.

See also

merge_expected_catch()