In DanOvando/GUM: Run Global Upsides Model

---

title: "GUM" author: "Dan Ovando" date: "r Sys.Date()" output: rmarkdown::html_vignette vignette: > %\VignetteIndexEntry{Run_GUM} %\VignetteEngine{knitr::rmarkdown} %\VignetteEncoding{UTF-8}

---

The GUM package runs the Global Upside Model presented in Costello et al. 2016. Once installed, you just need to pass it a data frame with at minimum variables IdOrig (id), Year (year), Catch (catch), and SpeciesCat, the numeric ISSCAAP species category. SciName (scientific name) helps as well. The rest should take care of itself.

Important: You must pass a data frame and NOT a data table as used with dplyr for the analysis to work. If you converted your data to a data table while preparing it be sure to convert it back to a true data frame using data.frame()

Additional parameters that can be supplied to the data frame as available are:

Variable | Description ----------|------------ MaxLength | The maximum length AgeMat | Age at Maturity VonBertK | Von Bert Growth Rate Temp | Preferred temperature b_to_k_ratio | Ratio of bmsy to K

Even if you only have these for selected stocks, the model will try and fill in missing values from FishBase.

First, let's take a look at some sample data

library(GUM)
library(tidyverse)
# devtools::load_all('GUM')
head(sample_data)

There's about 100 stocks in there, so let's subset this down to something smaller, and with less data to test the package

stocks = unique(sample_data$IdOrig)

sub = sample(stocks, 10, replace = F)

small_dat = filter(sample_data, IdOrig %in% sub)

less_dat = small_dat %>% dplyr::select(IdOrig,SciName,SpeciesCat,CommName,Year,Catch,BvBmsy) %>%
  mutate(IdOrig = as.character(IdOrig))

no_dat <- less_dat %>%
  mutate(SpeciesCat = 36, SciName = 'Blah')

results = run_gum_assessment(dat = less_dat)

results_no_sciname = run_gum_assessment(dat = no_dat) #test when no scientific name

ggplot(results,aes(MSY)) + 
  geom_histogram()

ggplot(results_no_sciname,aes(MSY)) + 
  geom_histogram()

a = ggKobe(dat = results)
a

DanOvando/GUM documentation built on May 6, 2019, 1:22 p.m.