SS3.36.16.00 released Sept 3, 2020 - Public Forums - Stock Synthesis

Latest SS News:  Sept 3, 2020

Greetings Stock Synthesis (SS) Users, 

This release of SS (3.30.16.00) includes both augmentations and bug fixes. There are no mandatory input changes. The most notable augmentations are:

• An option to reduce and customize the tables output in Report.sso.

• A size-based selectivity option to fix selectivity to be constant over a range of size bins.

• New parameter deviation and environmental linkage options to restrict the adjusted parameter within the low-high range of the base parameter (similar to a 3.24 option).

• SS is now compiled with ADMB 12.2 (instead of ADMB 12.0). This allows SS to work with the NUTS (No U-Turn Sampler) algorithm for MCMC (through ADMB command line options or the adnuts R package, which has a recent release).

We advise users to update their version of SS to take advantage of augmentations and bug fixes. Remember to update your copy of r4ss and SSI also as these work closely together. The full list of changes is in the change log, and an abbreviated list of changes since the last SS release is posted at the end of this message. Thank you to all users who reported bugs or requested features - your suggestions help us improve SS for everyone.

We have been improving on our SS build and test processes since the last release. We have implemented an automated build and test routine using Jenkins on a linux server, which runs about 25 complex stock assessment models using the latest code commit for SS and checks for changes in the model results. This allows us to run tests more frequently than before and hopefully results in catching bugs more quickly.

The latest 3.30.16.00 executables can be found in the SS document library.  Please note that between releases, fixes are posted in the latest executables folder.  A copy of the SS source code (the ADMB tpl) for this release is available to browse. 

GUI - Stock Synthesis Interface (SSI) updates: The GUI was updated to accommodate changes for this release of Stock Synthesis. For details, please see the SSI_readme.txt file that is included in the zipped GUI files. Have bug reports or feature requests for the GUI? Please send the SS team an email at nmfs.stock.synthesis@noaa.gov or post in the forums. 

ss3sim updates: The process to run simulations in ss3sim was simplified from needing several text files to the use of a single data frame; check out the vignette for an example. We hope to have the changes on CRAN soon.

r4ss updates: The biggest change is a rewrite of the SS_output() function (discussed on github) to use blank lines instead of the following keyword to mark the end of each table when parsing Report.sso. This was necessary to adapt to the potential absence of individual pieces due to the new customized reporting feature in SS. More information on changes to r4ss is available by looking at recently modified issues on the r4ss GitHub issues page.

Additional tools for SS updates:  An R package to use SS models in an MSE is under active development. Please visit the SSMSE repository for more information. Occasionally, we hear from users who have developed useful add-ons to SS. If you have such a tool, we’d love to hear about it. We can advertise it in SS release announcements or on the SS website if you would like and would be happy to consider whether a change to SS could make your package easier and better.

SS User Manual updates: An updated version of the user manual includes documentation of new features in the release. A citable version of the manual is available and can be cited as:

Methot, R. D., Jr., C. R. Wetzel, I. G. Taylor, and K. Doering. 2020. Stock Synthesis User Manual Version 3.30.15. U.S. Department of Commerce, NOAA Processed Report NMFS-NWFSC-PR-2020-05. https://doi.org/10.25923/5wpn-qt71

Contact us:  Please do not hesitate to report bugs,  ask a question about SS, or request a feature by posting on the forums or emailing the SS team at nmfs.stock.synthesis@noaa.gov. Please contact VLAB administrators at vlab.admin@noaa.gov to change your username or password or for other VLAB related questions. 

Rick Methot and the SS Team (alphabetically):

Kathryn Doering

Kelli Johnson

Neal Schindler

Ian Taylor

Chantel Wetzel

*********************************************

Change log since last release:

FEATURES

• Selectivity: add size selectivity option 11:  sel=1.0 over a bin range

• TimeVary:  creates new option to Z score or mean center environmental data within SS. Use this with a new environmental link option (#3), which uses a logit transform to constrain the result to the base parameter’s min-max. Value of the envlink should be constrained to about 1.8 to prevent U-shaped distribution of adjusted parameters.

• TimeVary:  create new parameter deviation approach (#5) that uses logit transformation to  constrain the result to the base parameter’s min-max.

• I/O: change init_equ_catch logL and lambda from scalar to fleet-specific vector

• I/O:  custom control of 60 reporting options; use report detail  option 3 in the starter.ss file to invoke custom reporting options

• I/O:  generate warning when accumulating catch specified more than once for the same yr, seas, and fleet

• I/O:  add another extra_std option to get se of dynamic Bzero

• I/O: add option in starter.ss allowing multi-year average of F_std

  • 11 # F_std_basis: 0=raw_annual_F; 1=F/Fspr; 2=F/Fmsy; 3=F/Fbtgt; where F means annual_F; values >=11 invoke multiyr with 10's digit

CORRECTIONS & REVISIONS

• Selectivity:  mirror, change length selex #5 to set selectivity = 1.0e-10 outside of range where selectivity = 1.0 to avoid gradient error; add more warnings for input of selex min-max ranges

• Benchmark: now initialize 2 vectors that occasionally caused erratic benchmark calculations

• Benchmark: input for the range of years to average now enforces input <= endyr

• Forecast: correct array index to allow forecast observations in a multi-area model

• Forecast:  correct indexing error for N-at-age in forecast when settlement age > 0

• Forecast:  correct timevary biology parameter indexing to operate in forecast

• Forecast: revise internal tracking to allow a fleet with no historical catch to operate in forecast (e.g., a forecast only fleet)

• I/O:  revise starter.ss_new and forecast.ss_new to display input values for year ranges in addition to the absolute years

• I/O:  fix output of RET and MORT in report.sso so it displays correctly for Benchmark and Forecast years; internal calcs were OK

• I/O: autogeneration of time-varying parameters: change ss_trans so that autogen set to 1 in control.ss_new in case the trans process has created some time-vary parameters