Skip to content

DTUAqua/spict

Repository files navigation

spict

An R-package for fittng surplus production models in continuous-time to fisheries catch data and biomass indices (either scientific or commercial). Main advantages of spict are:

  1. All estimated reference points (MSY, Fmsy, Bmsy) are reported with uncertainties.

  2. The model can be used for short-term forecasting and management strategy evaluation.

  3. The model is fully stochastic in that observation error is included in catch and index observations, and process error is included in fishing and stock dynamics.

  4. The model is formulated in continuous-time and can therefore incorporate arbitrarily sampled data.

Help files

A vignette for the package is available here, and serves as an introduction to the package and its functionality. The vignette also contains description of the more advanced features of the package.

A document with technical guidelines for using SPiCT is available here. This is a living document that has a list of things to check before accepting an assessment and some options to deal with more dificult data sets.

The package also contains reasonable documentation in the form of help texts associated with each function (some may not be fully up-to-date). These can be accessed in the usual R manner by writing e.g. ?check.inp. A good place to start (in addition to reading the vignette) is to read ?check.inp and ?fit.spict.

Citation

The underlying model used in the package is described in a published paper. A preprint of the paper is included in the package in the inst folder and can be downloaded here. To get citation information write citation(spict) in the command line.

Package requirements

The package requires TMB to be installed. TMB is now a part of CRAN and can therefore be installed using install.packages("TMB", type="source"). For more information about TMB click here.

Installing the spict package

There are two main options installing spict:

  • installing from the master branch that includes features that are thoroughly tested:
remotes::install_github("DTUAqua/spict/spict")               # master branch
  • installing from the development branch that includes more recent developments, and proposed bug fixes, but is less tested:
remotes::install_github("DTUAqua/spict/spict", ref = "dev")  # development branch

Windows

Installing spict in Windows requires Rtools, available here. When installing Rtools, it is important to check the option ""Add Rtools to the system PATH".

Then using the above procedure usinginstall_github() should work (make sure to remove spict before trying to reinstall).

If it doesn't work the old, but tedious, procedure can be used:

  1. Start 64 bit R and change working directory to the (cloned or unzipped) spict folder.

  2. From R run: source("install_windows.R")

When running install_windows.R remember to set your working directory to the spict directory containing install_windows.R.

Funding

The development of spict was partially funded by the European Maritime and Fisheries Fund (EMFF) and the Danish Fisheries Agency via the RoMA project (Grant agreement number 33113-B-20-183).