Release
notes for Stock Synthesis 3.30.19
April
4, 2022
Greetings
Stock Synthesis Users,
The release
of Stock Synthesis (3.30.19.00) is now available. There are
no mandatory input changes. We recommend that all users update to
the latest version of Stock Synthesis to take advantage of
augmentations and bug fixes. Recently, our team did a thorough
review of all accumulated requests and prioritized them for
development, so look forward to more advancements with 3.30.20 in
about 6 months. We’ll certainly be looking to integrate the newest
capabilities of ADMB.
The
most notable changes in 3.30.19 are:
-
The
data.ss_new file is now deprecated. Instead, data_echo.ss_new,
which echoes inputs to the data file with SS3 standardized
comments, is generated. If requested, the expected values will
be output into the file data_expval.ss, while each bootstrap
file will be output into a separate file called
data_boot_XXX.ss, where XXX is the number bootstrap. This change
was made because dividing up the files will make examining the
output files less error-prone. See the output
files section of the user manual for more information.
-
Optional
output to sub-folders. If the user creates a sub-folder named
ssnew, then all .ss_new files will go there (with their new
names). If the user creates a sub-folder named sso, then all
.sso output files will go there. We hope this new option will
ease management of output files.
-
A
new bioeconomic option created in collaboration with Andre Punt
has been added to benchmarks. This option allows the MSY search
to be in terms of value or retained catch, rather than dead
catch. It is still a beta feature. See the features
section of these release notes for how to invoke this option.
-
Added
a sub-option for cubic spline to control performance at tails.
-
Improved
predator M2.
-
Fixed
and improved output for hermaphroditic populations, in
particular for multi-season models and for male-to-female transitions.
-
Added
option to help work with tag-recapture data having small number
of returns
-
Deprecated
the helper spreadsheets. See the Selectivity
Web App and the User Manual instead.
For
Linux users, we have changed to compiling the Linux binary on CentOS
to Ubuntu. This is because we have easier access to Ubuntu machines
and the version of CentOS Linux we were using has reached
end-of-life. However, this may cause issues for some Linux users.
Please let us know through the forums or email if you are running
into difficulties due to this change in compiling. If necessary,
Stock Synthesis can be built
from source.
VLAB
updates: We are providing more resources on Github rather than on
the website. See the Stock Synthesis
homepage for the latest links.
GUI -
Stock Synthesis Interface (SSI) updates: The GUI
has been updated to accommodate changes for this release of Stock Synthesis.
ss3sim
updates: See the ss3sim
vignettes for information on getting started.
r4ss
updates: r4ss has been updated to maintain compatibility with Stock
Synthesis 3.30.19.
Additional
tools for SS updates:
-
The
helper spreadsheets have been deprecated, the information they
provided is available in other resources. Please see the user
manual for input file specifications and the SS3
helper shiny app or SSI for visualizations of the double
normal and logistic selectivity functions. Thanks to Allan Hicks
and Andrea Havron for developing the shiny app!
-
The
ss3diags
R package contains additional diagnostic routines for SS3 models.
SS3
User Manual updates: Most documentation for Stock Synthesis can be
viewed at https://nmfs-stock-synthesis.github.io/doc/.
In particular, there is now an html
version of the user manual.
Contact
us: Please do not hesitate to report bugs, ask a question about
SS, or request a feature by posting on the forums,
opening an issue
(for those with github accounts), or emailing the SS3 team at nmfs.stock.synthesis@noaa.gov.
Change
log since last release:
See Change_log_for_SS_3.30.xlsx
for more details.
Features
-
Revise
lines needed for West Coast rebuilder (new option 2 in forecast
file). Issue #276
-
Bioecon
benchmark beta feature - Benchmark options and output in
economic units. Issues #222,
#251
To use
this feature, use the following options
in the forecast file:
5 |
#
MSY Method (FMEY) | |
COND:
MSY Method = 5
1 |
Maximum
economic yield (MEY) units
1
= dead biomass;
2
= dead biomass without excluded bycatch fleet;
3
= retained biomass; and
4
= profits using price and costs. | |
1
0 0 1
-9999
0 0 0 |
MEY
options - Fleet, Cost/F, Price/F, and Include FMEY in Optimization |
To
calculate the FMEY enter fleet number, the cost per fishing
mortality, price per mt, and whether optimization should
adjust the fleet’s F or keep it at the mean from the
benchmark years (0 = no, 1= yes). Care should be taken when
scaling the values used for cost/F and price/mt. Units in
the example show cost=0 and price = 1, so will be identical
to MSY in weight. Note, if a fleet’s catch is excluded from
the FMEY search, its catch or profits are still included in
the MSY value using historical F levels from benchmark years |
-
Add
option for constant/0 selectivity below the first knot in cubic
spline. Issue #86.
See setup in the
cubic spline selectivity description of the User Manual.
-
Add
option to get a report file for each MCEVAL sample. Issue #70.
Evoke this by using this line in the starter file: 2 # MCMC
output detail
-
Allow
subfolders for .sso and .ssnew output; new filenames for ss_new
output. Issue #226.
Details available in the optional
output subfolders and the output
files sections of the user manual.
-
Augment
reporting of overall comps (table in Report.sso with the header
OVERALL_COMPS report:30) to inclue breakdown by season and sex.
Issue #215.
New output has these fields after adding area and season:
area
seas Fleet N_obs len_bins 10 14 18 22 etc.
New
rows include rows for season = 0 to report combined seasons when the
number of seasons is more than 1 and separate rows for males and
females in a two sex model.
Corrections
and Revisions
-
Fix
summary biomass sd so will work. Issue #271
-
Remove
deprecated selectivity patterns. Issue #272
-
Fix
length comp super-periods so that they work with retrospectives.
Issue #265
-
Fix
mirroring dispersion tag parameters. Issue #268
-
Fix
output for hermaphroditic populations. Issue #246
-
Improve
weight at age storage efficiency. Issue #136
-
Revise
F method 4 to allow constant F. Issue #243
-
Fix
multi-season forecast catch output. Issue #205
-
Fixes
for predator M2. Issue #45
-
Add
option for dealing with tag-recapture data having low N
recaptures #280
-
convert
catchfleet IF statements to a lookup table. Issue #192