Data

Tools

Indexes

Applications

Experiments

Open Source Science

Recent Releases of t3

t3 - t3: A R package for processing tropical tuna fisheries data

t3 3.0.4 - 2025-07-04

Added

  • Improve documentation :
    • Add reference table's documentation and easy access :data(package="t3").
    • Add documentation about outputs in ".csv"" format of function full_trips$rf1().
    • Add "elementarycatchid", "statutrf1label", "statutrf2label" columns, in ".csv" output named "process11detail", of function full_trips$rf1().

t3 3.0.3 - 2025-06-12

Added

  • Add the argument applyrf1onbycatch=TRUE/FALSE, in the method full_trips$rf1() By default TRUE, the raising factor value, calculated for each trip, is applied to all the logbook catches associated to the trip, including by-catch species. If FALSE, only the catch weights of species belonging to the species list defined by the `speciesfaocodesrf1` argument are corrected, rf1 is not applied to by-catch species.

Changed

Removed

  • Remove argument output_format="eu"/"us" to uniformize output format with sep="," and dec=".".

t3 3.0.2 - 2025-05-21

Changed

t3 3.0.1 - 2025-05-13

Changed

  • Important calculation time reduction by replacing future.apply::future_sapply() by sapply() in method extract().
  • Fix typo (weight instead of weigth) in "inst/avdth_samples.sql".
  • In method fishing_effort():
    • For set declared as positive but with no elementary catch associated, the set duration is now set to the null set duration value, taken from the “inst/setdurationref.csv” table, and not to NA as before and a warning is generated.
    • If there is no adjusted set duration model for the year corresponding to the activity date in the reference table (“inst/setdurationref.csv”), t3 now allows considering the adjusted set duration model for the year closest to the activity date in the reference table, generating a warning.

Added

  • Add global_output_path argument in function pathtolevel3 to save the file named "inputslevl3target-yearoceanocean-codeflag-codes.Rdata", in outputs folder, if `globaloutput_path` is not NULL.

t3 3.0.0 - 2025-04-15

Changed

Added

  • Add global_output_path argument in function pathtolevel3 to save inputslevl3target-yearoceanocean-codeflag-codes.Rdata, in outputs folder, if `globaloutput_path` is not NULL.
  • Add method fishing_effort() gathering methods 1.5 to 1.8:
  • setduration : calculated according to a linear function of catch weight with two parameters a and b. These are found through a reference table (setduration_ref.csv), for each year, ocean, fishing school and country.
  • timeatsea : the process divides the day's time at sea declared between the activities, allowing the allocation of time at sea, recorded on that date. If no activity to allocate time at sea is recorded on a given date, with a non-zero time at sea, a transit activity is created (whose id_activity contains #666#) to allocate the time ate sea of that date.
  • fishingtime : the process module the duration of a working day according to the real sunrise and sunset of each day. It then divides the day's fishing time between the fishing activities recorded on that date. If no fishing activity is recorded on a given date with a non-zero fishing time, a searching activity is created (whose idactivity contains #666#) to allocate the fishing time of that date.
    • searchingtime = fishingtime - set_duration.

Removed

  • Methods 1.5 to 1.8 and 1.2 (rf2).

Warning: you will need to correct this in your scripts

Full Changelog: https://github.com/OB7-IRD/t3/compare/2.1.2...3.0.4

- R
Published by JeanneClement 8 months ago

t3 - t3: A R package for processing tropical tuna fisheries data

t3 2.1.2 - 2025-04-14

Changed

Added

  • Possibility of considering AVDTH and Observe data as input for model fitting at level 3.

Full Changelog: https://github.com/OB7-IRD/t3/compare/2.1.1...2.1.2

- R
Published by JeanneClement 11 months ago

t3 - t3: A R package for processing tropical tuna fisheries data

t3 2.1.1 - 2025-03-21

Changed

  • Manage fishing effort calculation for activities with multiple floating objects or object operation code declared.

Full Changelog: https://github.com/OB7-IRD/t3/compare/2.1.0...2.1.1

- R
Published by JeanneClement 11 months ago

t3 - t3: A R package for processing tropical tuna fisheries data

t3 2.0.1 - 2025-01-21

Changed

  • Change the type of the flag_codes parameter of object_model_data() function from integer to character (three-letter FAO code(s) for the country(ies)).

t3 2.0.0 - 2024-11-29

Added

  • Add data source observe_database.
  • Add weight category + 60kg (code 14).
  • Add functionality for querying multiple databases, for example the main and acquisition observe databases to simultaneously import and process ‘recent’ data from acquisition database, not yet imported into the main database, and older data from the main database.
  • Add activity code referential to allocate time at sea and fishing time and set duration.

Changed

  • Update vignettes.
  • Update unit tests.
  • Use of codama for checking arguments.
  • Update referential tables : setdurationref, lengthstep and lengthweight_relationship.

Removed

  • Remove data source t3_db.

Full Changelog: https://github.com/OB7-IRD/t3/compare/1.1.0...2.0.1

- R
Published by JeanneClement 11 months ago

t3 - t3: A R package for processing tropical tuna fisheries data

t3 2.1.0 - 2025-03-19

Added

Changed

  • Optimization :
  • Update the activity code referential for time allocation (timeallocationactivitycoderef.csv)
  • Improve fishing effort evaluation. The duration of activities is now calculated according to the activity code referential, in order to allocate time only to significant activities:
    • setduration : calculated according to a linear function of catch weight with two parameters a and b. These are found through a reference table (setduration_ref.csv), for each year, ocean, fishing school and country.
    • timeatsea : the process divides the day's time at sea declared between the activities, allowing the allocation of time at sea, recorded on that date. If no activity to allocate time at sea is recorded on a given date, with a non-zero time at sea, a transit activity is created (whose id_activity contains #666#) to allocate the time ate sea of that date.
    • fishingtime : the process module the duration of a working day according to the real sunrise and sunset of each day. It then divides the day's fishing time between the fishing activities recorded on that date. If no fishing activity is recorded on a given date with a non-zero fishing time, a searching activity is created (whose idactivity contains #666#) to allocate the fishing time of that date.
    • searchingtime = fishingtime - set_duration.

Removed

Full Changelog: https://github.com/OB7-IRD/t3/compare/2.0.1...2.1.0

- R
Published by JeanneClement 11 months ago

t3 - t3: A R package for processing tropical tuna fisheries data

t3 2.0.1 - 2025-01-21

Changed

  • Change the type of the flag_codes parameter of object_model_data() function from integer to character (three-letter FAO code(s) for the country(ies)).

t3 2.0.0 - 2024-11-29

Added

  • Add data source observe_database.
  • Add weight category + 60kg (code 14).
  • Add functionality for querying multiple databases, for example the main and acquisition observe databases to simultaneously import and process ‘recent’ data from acquisition database, not yet imported into the main database, and older data from the main database.
  • Add activity code referential to allocate time at sea and fishing time and set duration.

Changed

  • Update vignettes.
  • Update unit tests.
  • Use of codama for checking arguments.
  • Update referential tables : setdurationref, lengthstep and lengthweight_relationship.

Removed

  • Remove data source t3_db.

What's Changed

  • 135 détermination du mode de pêche cas particulier dabsence de données pour le niveau 34 by @Antoine-Duparc in https://github.com/OB7-IRD/t3/pull/136
  • 116 add target flag in level 3 and option of sample sharing among fleet by @Antoine-Duparc in https://github.com/OB7-IRD/t3/pull/139

Full Changelog: https://github.com/OB7-IRD/t3/compare/1.1.0...2.0.1

- R
Published by JeanneClement 12 months ago

t3 - t3: A R package for processing tropical tuna fisheries data

The multi-species nature of tropical tuna surface fisheries highlights a series of difficulties when catches are estimated by species and size statistics. The Tropical Tuna Treatment (T3) process was initiated more than 30 years ago and is still today the result of a strong collaboration between several countries (notably France, Spain, the Seychelles and Senegal). Its aims are to correct biases of the logbook data on species composition and provide more accurate catch estimations per species for the purse seiner fleet, especially European and Seychelles fleet.

The process is composed of three levels : - The standardisation of the logbook data - The standardisation of the sampling data
- The modelling of species composition and catch estimations

There is two output :
- Catch estimates: annual catch by species, region, gear, flag. - Fishing effort: catch and fishing effort for each species by small area (1x1 degree squares), gear, flag and month.

Added features:

  • Workflow is full completed for AVDTH

What's Changed

  • Fix renaming 3.1 by @Antoine-Duparc in https://github.com/OB7-IRD/t3/pull/68
  • fix output directory issue 3.3 by @Antoine-Duparc in https://github.com/OB7-IRD/t3/pull/69
  • Fix format level 3 issues by @Antoine-Duparc in https://github.com/OB7-IRD/t3/pull/70
  • Update full_trips level 3 process 3.3 by @Antoine-Duparc in https://github.com/OB7-IRD/t3/pull/71
  • Dev l3 20230316 by @Antoine-Duparc in https://github.com/OB7-IRD/t3/pull/72
  • Dev level3 by @Antoine-Duparc in https://github.com/OB7-IRD/t3/pull/96
  • Dev level3 by @Antoine-Duparc in https://github.com/OB7-IRD/t3/pull/100
  • Development by @Antoine-Duparc in https://github.com/OB7-IRD/t3/pull/101
  • Dev level3 by @Antoine-Duparc in https://github.com/OB7-IRD/t3/pull/106
  • Multi year run issue by @Antoine-Duparc in https://github.com/OB7-IRD/t3/pull/108
  • fix missing weight cat LB by @Antoine-Duparc in https://github.com/OB7-IRD/t3/pull/110
  • add process_level3 name by @Antoine-Duparc in https://github.com/OB7-IRD/t3/pull/113
  • Fix LTA issue #114 by @Antoine-Duparc in https://github.com/OB7-IRD/t3/pull/118

Full Changelog: https://github.com/OB7-IRD/t3/compare/1.0.0...1.1.0

- R
Published by JulienLebranchu over 1 year ago

t3 - t3: A R package for processing tropical tuna fisheries data

The multi-species nature of tropical tuna surface fisheries highlights a series of difficulties when catches are estimated by species and size statistics. The Tropical Tuna Treatment (T3) process was initiated more than 30 years ago and is still today the result of a strong collaboration between several countries (notably France, Spain, the Seychelles and Senegal). Its aims are to correct biases of the logbook data on species composition and provide more accurate catch estimations per species for the purse seiner fleet, especially European and Seychelles fleet.

The process is composed of three levels : - The standardisation of the logbook data - The standardisation of the sampling data
- The modelling of species composition and catch estimations

There is two output :
- Catch estimates: annual catch by species, region, gear, flag. - Fishing effort: catch and fishing effort for each species by small area (1x1 degree squares), gear, flag and month.

Added features: - Implementation of unit tests - Implementation of the documentation for the level 3 - Development of output extraction functions for process 1.1 - Adding an output directory intilisation function and integrating it into the processes

Changed features: - Fix minor bugs and optimisation of the code and regarding the overall process - Update of the level 3 process

Full Changelog: https://github.com/OB7-IRD/t3/compare/0.9.1...1.0.0

- R
Published by JulienLebranchu almost 4 years ago

t3 - Beta version of T3 software

Added

  • Implementation for the import of annual dataset from outputs of the level 1 and 2 (period length customizable)
  • Implementation of boostrap method as base for all confidence intervals
  • Computation of confidence interval (bootstrap interval) for the nominal catches by species and by fishing mode
  • Implementation of outputs for the he nominal catches by species and by fishing mode

Changed

  • Improve modelling configuration (addition of new parameters)
  • Rewrite several figures for the model checking
  • Improve time computing using faster implementation of random forests for the species composition modeling in the level3

Removed

  • Disabling task and effort computation while waiting for confident interval implementation

- R
Published by JulienLebranchu over 5 years ago

t3 - Beta version of T3 software

Beta release of T3R package

This version contains a functional workflow to process tuna data.

Major functionalities added : - Implementation of the model layer in R6class - Implementation of all logbooks standardization sub-processes - Implementation of all samples standardization sub-processes - Implementation of sub-processes for assessing species composition in relation to non-sampled assemblages - Strict selection of samples used in the model training according to several criteria on sets and wells: school type, location, date, sample quality.
- Classification of unknown school type (no data available) by clustering according to major tuna species composition.
- Development of a statistical model accounting for: - Spatial and temporal structure of the fishery: location and date of the catch - Vessel specificity which impacts the species composition of the catch - Crew reporting information which enables more flexibility in the model when a marginal composition is caught - Several years of fishing in the training to increase the robustness of the estimates - Model checking to ensure reliability of the results - Computation of a confidence interval (bootstrap method) which give the trust level on catch estimations.

Note: The functionality to import data from RData or CSV files is under development.

- R
Published by JulienLebranchu over 5 years ago