- Added season to the time series results and stopped hard-coding to season 1.
- Added month to the time series results.
- Use better semantic versioning protocols where the minor number is reserved for features and the patch number (terminal) is for fixes.
- Fixed bug in Introduction vignette regarding self-test estimation method to allow for changing multiple parameters
- Updates {ss3sim} to use Stock Synthesis v3.30.20.
- Removes
shQuote()
fromget_bin()
where file paths with spaces are no longer quoted.
- Updates {ss3sim} to use {r4ss} version 1.46.1
tune_comps()
run()
, which requires removal ofhess_always
call and instead usesextras
call torun()
to turn the hessian off.
- Depends on {stringi}.
- Removes deprecated arguments in
change_e()
. - Adds
sds_out
tosample_index()
thanks to the request and review by Leire Citores. Now, bias in the estimation method regarding the uncertainty of the index can be tested.
- Updates
change_q()
to more modern code and removes bugs. Errors are no longer generated when the fleet does not exist. find_position()
simplifies the code to find the position of a fleet name in a vector. As well as, the code was moved to an exported function for others to use outside ofchange_q()
.
- Bump version to match current version of SS3, where version numbers of {ss3sim} includes three parts separated by full stops. The first number represents the major version of {ss3sim}. The second number is the minor version, i.e., third set of digits, of the SS3 executable used in {ss3sim}. The third number is the minor version of {ss3sim}.
- Remove dependency on {bbmle} to externally estimate growth.
- Deprecate functions that were not being used in the code base, i.e.,
cleanup_ss3()
,get_bin_info()
,change_rec_devs_par()
, ... - Does not turn on empirical weight-at-age in EM if sampling weight-at-age data. Instead, users must specify to use empirical weight-at-age data in their control file for the estimation method allowing users to sample the data but not use it if they do not want to.
- Allows for no estimation method via
em_dir = NA
, creating a way to simulate data quickly.
- Fix bug related to tidyverse in
sample_comp()
where the composition data were appended to the data frame as a matrix rather than integrated as additional columns.
- Deprecate the use of the log file in
ss3sim_base()
. The argumentss3sim_base(print_logfile = FALSE)
has a new default and will warn users if they have the value set toTRUE
. Fully deprecated in {ss3sim} version 1.19.3. - Create data objects for the OM and EM control and data files. These files are often used in examples.
- Deprecate
change_f_par()
which was not being used by any downstream code because {ss3sim} uses the control file instead of the par file. - Move Anderson et al. (2014) to a dedicated repository ss3sim/ss3sim_andersonetal.
- Allow for seas and partition in sampling.
- Drop fleets from sampling regimes using
NA
in the scenario data frame. - Upgrade to Stock Synthesis 3.30.16.
- Start allowing for seasons in data.
- Fixed bug in how q was being checked.
- Partition
change_catch()
fromchange_data()
. - Allow for fleet specific
cpar
andESS
values. - Allow for no SPR in results file.
- Fix bad param argument in
weight_comps()
with DM.
- Control height and width of jittering in
plot_points()
withjitter.height
andjitter.width
.
- Allow for custom name of results files.
- Upgrades to Stock Synthesis 3.30.15.03.
- Deprecates case files in favor of
run_ss3sim(simdf =)
. - Uses more templates for parameter arguments, e.g.,
par_int
. - Simplifies scenarios, iterations, OM/EM to
dir
to increase the usability of functions outside of {ss3sim}. - Adds
type.convert(as.is = TRUE)
because of R 4.0. - Uses maxfn 0 to increase speed of tests and run more tests on CRAN check.
- Figures
- Uses better labels for x and y axes in
ggplot2::ggplots()
; - Changes RE line to dashed black;
- Passes coloration of lines in a better way;
- Uses
plot_ss3sim()
as back-end code for plotting functions; - Uses character strings as input arguments with calls to data;
- Adds cumulative mean figure; and
- Allows for fill in boxplots;.
- Uses better labels for x and y axes in
- Results:
- Creates long-data format for results to allow functions to be used by {SSMSE};
- Removes columns that are not needed;
- Removes row names in summary files;
- Renames rec_dev to increase clarity;
- Creates better grep of management quantities;
- Removes parallel capacity;
- Only reads comp file if present by changing CompFile to
r4ss::SS_output()
; - Allows for relative directories for {SSMSE}.
- Allows for multiple fisheries in
change_f()
and change fisheries to fleets. - Increases robustness of
get_success()
. - Break tests into multiple chunks to pass Travis-ci checks.
- Allows weighting of composition data.
- Starts EM in first year with non-zero catches.
- Reinstate bias adjustment, but once for every iteration.
- Add
create_logo()
to generate GitHub logo. - change travis configuration to use Xenial.
- Realized broken links are because of case not full path and changed all links back to shorter path with lower-case first letter.
- Includes NOAA disclaimer in README.
- Changes link in inst/doc/making-models.html to a more explicit link using ../doc/introduction.html.
- Fixes broken links in README.md and inst/doc/introduction.html.
- Confirmed that the links in inst/doc/making-models.html are working.
- Improves documentation.
- Updates, adds, or removes examples.
- Updates
profile_fmsy()
to allow for relative or absolute paths to directories and provide a better message ifverbose = TRUE
when users may have not run their model long enough to reach equilibrium. - Removes reliance on {tidyverse}.
- Changes to 'iteration' instead of 'replicate'.
- Updates the license to 2019.
- Changes README from Rmd to md file.
- Deletes the vignette on making functions and updated remaining vignettes.
- Removes bias adjustment capability and sampling of conditional age-at-length data; both will eventually be added back to the package with better code
- Removes
write_file
argument from many functions and rely on other input arguments to determine if a file should be written to the disk, such as if the name of the file isNULL
then nothing is written. - Creates
change_o()
for the OM so users can input INIT values for parameters in their OM rather than having to use a dev vector andchange_tv()
. - Implements check of q parameter such that a given q is created from a template for each survey-like data set included after sampling.
- Manipulate control file rather than par file and delete the par file from the OM folder.
- Use
merge()
rather thancbind()
for those instances that the names of the columns in the results file from the OM do not exactly match the EM. - Fixes
get_results()
working directory issue. - Deprecates
change_year()
,change_fltname()
,change_maturity()
, andpastef()
.
- Reduces restrictions placed on the OMs and EMs used by individual users to allow for multiple fleets, two-sexed models, and other trivial changes such as different fleet names.
- Implements the use of more {r4ss} functions.
- Increases the number of functions with examples and update examples to not change the users working directory when possible.
- Provides a single model inside the package rather than maintaining models both in the package and in ss3sim/ss3models.
- Kelli F. Johnson is now the maintainer of the package.
- Updates authors to reflect current contributors.
- Reduces the complexity of sampling functions by initiating the same protocol for ages and lengths, where each function calls a simple function.
- Updates functions to work with latest version of {r4ss}, which is compatible with Stock Synthesis 3.3.0
- Removes option to specify safe or optimized version of SS3.
- Adds citation to Description.
- Fixes compatibility with {dplyr} 0.6.0.
- Fixes issue with SS3 binary in development version #239.
- Fixes LICENSE for CRAN.
- Updates author list.
- Incorporate @cstawitz's additions to the
making-models
vignette. - Fixes to executable locating for Windows.
- Updates DESCRIPTION.
- Fixes compatibility with {dplyr} 0.4.3.9001.
- Fixes compatibility of vignette with {ggplot2} 2.0.0.
- Adds new data types:
- conditional age at length,
- mean length at age, and
- empirical weight at age.
sample_agecomp()
andsample_lcomp()
now take new effective sample size (ESS) argument as inputs, so the user can weight the composition data as desired.- Add functions to estimate von Bertalanffy growth. These are used internally
by
change_e()
. - Add
change_maturity()
to alter the SS3 maturity option. - Check that fleet name in the data frame matches the newest SS3 version (see
change_fltname()
). - Add
change_data()
, which prepares the OM .dat files to have the correct data used for sampling that scenario. It also can manipulate length bin widths, the SS3 robustification constant, and the SS3 tail compression value. It takes care of calling the newchange_tail_compression()
andchange_lcomp_constant()
functions. - Add
calculate_re()
to calculate relative errors on the result data frames. - Add internal checking of the validity of SS3 .dat files (#203).
- The
get_results()
function now returns the following:- run time,
- a list of parameters that were on bounds, and
- recruitment deviations.
- Switches vignette to R Markdown and HTML output (#194).
- Adds re-binning within the EM for length and conditional age-at-length data (#201, #205).
- Add Kelli's
profile_fmsy()
function (#173). - Merrill added parallel processing option to the get results functions (#171).
- Adds parallel iterations processing option (#168).
- Function argument descriptions now indicate which ones need to be specified in case files with an asterisk (#161).
- Merrill updated the get-results functions to work without bias runs, to check for folders that look like ss3sim runs, and to issue a warning if there are problems reading the files (#155).
- Copy wtatage.ss_new from OM to EM (#163).
- Adds Travis CI testing of the package https://travis-ci.org/ss3sim/ss3sim.
- Adds much more extensive testing of the package in the root
tests
folder. This is for internal testing only and not bundled with the package (#156). - Allow for missing arguments in case files. This allows us to add new
arguments to existing functions without breaking old code. A missing argument
gets passed as
NULL
via theadd_nulls()
function (within #148). - Re-write case file parsing to be more robust and allow for scientific notations (#157).
- Adds the option to use the optimized Stock Synthesis executables (#138 and #147) and ability to have spaces in the path name to the executable.
- Adds Cole Monnahan's plotting functions (#134).
change_lcomp()
,change_agecomp()
, andchange_index()
are nowsample_lcomp()
,sample_agecomp()
, andsample_index()
(#145).- Requires the SS3 binary to be named according to the version (#138 and #147).
- Bump required version of {r4ss} to 1.22.1.
- Requires R 3.1.0.
- Adds ability to pass seed value to
get_recdevs()
(#125). - Adds dimension check of
user_recdevs
to ensure recruitment deviations are supplied for all iterations (#130). - Uses a custom Solarized color theme in vignette.
- Moves vignette data files to the
data
folder so they are user-accessible after installing the package (#126). - Updates out-dated paper citations. All papers are now accepted.
- Removes multi-core code in vignette for CRAN checks.