Skip to content

Latest commit

 

History

History
75 lines (47 loc) · 5.5 KB

update-steps.md

File metadata and controls

75 lines (47 loc) · 5.5 KB

This repository contains:

  • a script to go from raw data (MCMC samples usually; stored locally) to cleaned and processed data (data-raw-process/process-raw-data.R)
  • the processed data sets themselves as .rds files (data-generated)
  • the model in Stan analysis/rw-ss.stan
  • a simulation test of the model (analysis)
  • code to fit the model and make figures (analysis)
  • text for the preprint (preprint)

A. Adding and processing the latest assessment data

  1. Use the data-raw/process-raw-dat.R file to clean any newly added/updated assessment data (raw data can be placed in data-raw/model-output). This script saves outputs as: data-raw/species-region-mcmc.rds, where species name is often abbreviated

If the LRP and USR are based on fractions of Bmsy, process new data to get dataframe of the format:

year b bmsy run lrp usr species region iter
2018 1370 523 1 209 418 pacific-ocean-perch 3CD 1
2019 1400 523 1 209 418 pacific-ocean-perch 3CD 1
2020 1530 523 1 209 418 pacific-ocean-perch 3CD 1

If the LRP and USR are not based on fractions of Bmsy, the dataframe will look like this:

year blrp busr iter species region run
1918 7.26 3.63 1 quillback 4B 1
1918 9.00 4.50 2 quillback 4B 1
1918 8.34 4.17 3 quillback 4B 1
  • If there is no run column, you will need to add a placeholder, e.g., mutate(run = 1)

  1. Add any new stocks (if any) to the analysis/stock_df.R and data-raw/species-regions-tofit.csv (see Section C)

  2. Update the data-raw/last-assess-years.csv, grab the info from the latest stock assessments/stock assessment drafts in review.

B. Analyse the assessment data {#section-B}

The scripts contained in analysis summarise the assessment data. The following scripts should be run in order (as numbered) and should not need to be modified year to year. With the exception of updating end_year in analysis/01-stitch-data.R L.5 and analysis/stock_df.R.

  1. analysis/01-stitch-data.R can be run once the mcmc data files have been updated, collates the processed assessment data. UPDATE THE end_year on L5. This script also outputs a figure that you can use to check that all the B status ratios loaded properly and were updated properly before you fit the Stan model.

  2. analysis/02-fit-models.R compiles and fits the Stan model.

  3. analysis/03-plot-models.R plots model fits of the assessment time-series.

  4. analysis/04-ridges.R plots posterior distributions of estimated biomass over limit reference points

  5. analysis/05-combine-plots --> archive

  6. analysis/06-values.R --> archive

C. Analyse the survey data {#section-C}

The scripts contained in analysis-survey summarise the survey data. The following scripts should be run in order (as numbered) and should not need to be modified year to year.

If new species or assessment regions are added, you will need to update: data-raw/species-regions-tofit.csv and analysis-survey/get-all-spp.R.

  1. analysis-survey/00-get-all-spp.R is run to get the latest data on all the species of interest. This list may or may not need to be updated. Note this queries GFBio and will need to access to the DFO network.

  2. analysis-survey/02-render-indices.R

    • analysis-survey/01-index-new-deltas.Rmd is not manually run. It generates html files for each species and fits the delta-gamma model. If the delta-gamma fails to converge, the Tweedie is then fit. In 2023 all delta-gamma models converged.
    • analysis-survey/00-make-grids.R is also sourced by 02-render-indices.R. If prediction grids are updated this can be done here.

  3. analysis-survey/03-calc-slopes.R calculates the slopes of indices since 2000 so they can be ordered in Figure 3.

  4. analysis-survey/04-join.R joins the assessment data with survey indices and scales by the geometric mean.

  5. analysis-survey/05-plot-assess-surveys.R plots the latest assessments overlaid with the survey indices by species-region.

  6. analysis-survey/06-sopo-text.Rmd can be rendered to get an updated draft. Values used in the text are calculated at the top of the document, use those and manually update last year's word document (update with what SOPO sends as the template/version from the previous year), or choose to use the markdown document. Currently, references are not included and are added manually.

  7. analysis-survey/06-plot-presentation-figs.R generates the figures for the presentation.