README.md

In pbs-assess/hookCompetition: An R package for estimating relative abundance indices from longline data using a censored likelihood approach

hookCompetition

An R package for estimating relative abundance indices from longline data using a censored likelihood approach

Installation

remotes::install_github('pbs-assess/hookCompetition', dep=T, build_vignettes = TRUE)

Notes on how to use on non-IPHC data

To use the package on custom data, bypass the read_Data_hookcomp() step and ensure the following variable names are used inside the sf points object containing the catch counts:

• N_it_species (where species is the name of the species used (e.g. yelloweye)). This contains the catch counts of the species
• effSkateIPHC contains a measure of effort (e.g. number of hooks used etc.,)
• prop_removed contains the proportion of baits removed from the fishing event
• year the year of the fishing event
• region_INLA the region where the fishing event took place. An index (numeric variable) matching the corresponding row in survey_boundaries
• station containing the index of the fishing station where the fishing event took place. Useful if the same locations are visited each year (for multiple years)
• twentyhooks a binary variable indicating years where a different number of hooks were used. Useful if 2 fishing protocols are used and differences want to be controlled for. Set to 0 if not relevant.

Simulation Study

The simulation study accompanying the paper can be found in the folder 'Simulation Study'

Notes from Joe when Andy visited him in Sept 2022 and by Zoom in Oct 2022

Sean and Jillian -- some of this should be useful. I haven't had a chance to run anything more. I think for IPHC data it will use data as saved by my gfiphc package, which I haven't updated yet with 2022 data.

Joe pushed new censored_index_fun_sdmTMB() to hookCompetition, for operationally making indices.

Zoom chat 24/10/22 - IPHC data suggests fitting unstructured (no spatial) models with unique fixed effects for each year to get a coastwide index, plus spatiotemporal so can match spatial maps (see above function I think).

For generating new indices see Example.Rmd vignette, but for proper use (?) recommend using the functions that end in ...sdmTMB() to go into pacea, for example.

Use Censored_Longline_RCode (joe's GitHub) as submitted with manuscript, if have to re-make any figures. .rds files of results are saved. hookCompetition/Case Study is original code that then got tidied up into Censored_Longline_RCode/

Andy: ran hookCompetition/vignette/Example.Rmd at PSEC but was slow (think just due to connection): looks like code extracts new .rds files, but Andy ran from vignettes/ where it's saved them. They should match the ones sent by Joe in ../, maybe just move those to dummy folder so I have them, but I think they should be duplicated.

pbs-assess/hookCompetition documentation built on April 27, 2023, 12:47 p.m.