Skip to content

Latest commit

 

History

History
88 lines (65 loc) · 1.86 KB

repository_types.md

File metadata and controls

88 lines (65 loc) · 1.86 KB
Raw

TAF stock assessment repository types

A TAF stock assesment repository can be for several purposes. The we the purpose is identified is through a variable called <purpose> and is used in the repository name:

<year>_<stock code>_<purpose>

for example

2019_cod.27.47d20_forecasts

The list of <purpose>s are as follows (to be created as vocabularies):

  • data
  • data.survey
  • data.catch
  • assmnt
  • forecasts
  • refpoints

The data type can be split down into the type of data analysis being done:

  • data.survey
  • data.catch

there is also the possibility of using a type for benchmark assessments

  • benchmark

which an be split down and use the other types

  • benchmark.<type>

for example:

2019_cod.27.47d20_refpoints
2019_cod.27.47d20_data.survey
2019_cod.27.47d20_benchmark.data.survey

this last examlple would be a TAF repository to hold the data preparation scripts for creating the survey index developed at a benchmark workshop.

adding alternative assessments and procedures

In the case of a benchmark it is often of interest to run multiple assessments on the same data. This can be accomodated by creating a seperate TAF repo for each assessment, using a extended nameing convention

<year>_<stock code>_<purpose>_<identifier>

for example

2019_cod.27.47d20_assmnt_xsa

or

2019_cod.27.47d20_assmnt_multifleet

The <identifier> type for now is less strictly controlled, but could be linked (loosely) to a set of vocabularies, such as the model name https://vocab.ices.dk/?ref=1525, survey name vocabulary?

As JSON

as a JSON structure for

2019_cod.27.47d20_benchmark.data.survey_ibtsq1

would be something like:

{
  "year": "2019",
  "stock": {
    "stockcode": "cod",
    "areatype": "27",
    "areacode": "47d20"
  },
  "purpose": "benchmark.data.survey",
  "identifier": "ibtsq1"
}