Creating taf datasets - ices-taf/doc GitHub Wiki

See also: Bib entries, Example data records.

This page was based on using the icesTAF package version 4.2.0 dated 2023-03-21.

In this guide

• Clone
• Make Skeleton
• Upload initial data
  • Upload files
  • Make data available

Clone

The first step is asking the ICES secratariat to create a repository for your stock. In this example we will work with North Sea cod: 2019_cod.27.47d20_assessment in 2019 - the repository name for the assessment will be 2019_cod.27.47d20_assessment. When a new repository is created on GitHub you should clone the repository to your own computer. From here we will assume you have cloned the repository.

Make skeleton

The first step in creating a TAF analysis is to set out the basic folder and file structure of the project. This is done using the function taf.skeleton, which creates the following structure in your working directory

                             
 2019_cod.27.47d20_assessment
  ¦--boot                    
  ¦   °--initial             
  ¦       °--data            
  ¦--data.R                  
  ¦--model.R                 
  ¦--output.R                
  °--report.R                

Upload initial data

The next step is to set up the data requirements for your assessment. There are three ways to get data into an assessment: (1) upload files directly, (2) get a file from the web, and (3) use R code to access a web service. We will cover 1 and 2 here. See Example data records for more information.

Upload files

To upload a file called catch.csv - the contents might be something like:

year,catches
2015,2109
2016,3455
2017,3466
2018,2050

Simply copy the file in to the boot/initial/data folder. The directory structure will now look like this:

                             
 2019_cod.27.47d20_assessment
  ¦--boot                    
  ¦   °--initial             
  ¦       °--data            
  ¦           °--catch.csv   
  ¦--data.R                  
  ¦--model.R                 
  ¦--output.R                
  °--report.R                

Make data available to assessment

So far we have uploaded data into the intial data folder, but to make it available to the assessment it needs to be in the boot/data folder, and the only way to do this is to create a file reference and then run the function taf.boot. To reference the data we create a file called DATA.bib using the helper function draft.data, and add a few fields:

draft.data(
  originator = "WGNSSK",
  year = 2019,
  title = "Catch data for cod.27.47d20",
  period = "2015-2018"
)

## @Misc{catch.csv,
##   originator = {WGNSSK},
##   year       = {2019},
##   title      = {Catch data for cod.27.47d20},
##   period     = {2015-2018},
##   access     = {Public},
##   source     = {file},
## }

Then we write this to the DATA.bib file by specifying file = TRUE:

draft.data(
  originator = "WGNSSK",
  year = 2019,
  title = "Catch data for cod.27.47d20",
  period = "2015-2018",
  file = TRUE
)

Finally we run taf.boot() to get the data into the boot/data folder, leaving the directory tree looking like this:

                             
 2019_cod.27.47d20_assessment
  ¦--boot                    
  ¦   ¦--data                
  ¦   ¦   °--catch.csv       
  ¦   ¦--DATA.bib            
  ¦   °--initial             
  ¦       °--data            
  ¦           °--catch.csv   
  ¦--data.R                  
  ¦--model.R                 
  ¦--output.R                
  °--report.R