The latest release of the MSEtool package is available on CRAN.
data$I_type is now obsolete to remove redundancy with argument s_selectivity. Use s_selectivity to specify the selectivity of surveys. If data$I_type is detected, the code will attempt to update s_selectivity.data$MS (mean size) is now used instead of data$ML (fishery mean lengths). data$MS can also be mean weights, specify with data$MS_type to be either “length” or “weight”. The likelihood of data$MS uses a normal distribution with constant CV specified in data$MS_cv (default = 0.2).data argument can now be either a list (preferred) or a DLMtool Data S4 object.CAL_mids and Vmaxlen.Data@AddInd.AddInd calcs for DLMtool 5.4.4.SRA_scope can be set up with blocks. Unique blocks are defined and then assigned to fleet and year. New vignettes and updated help files for SRA_scope describe the set up in the function call.SSS, catch-only method with fixed depletion assumption) is now added to the package.DD_TMB and DD_SS) can now be conditioned on catch (previous versions only allowed conditioning on effort). The default is now to condition on catch, which is standard practice.SP and SP_SS now support multiple indices in the model, using Data@AddInd and Data@CV_AddInd. These assessments still support Data@Ind but a custom wrapper function is still needed to use either Data@SpInd or Data@VInd.SRA_scope has been added.SRA_scope can now solve F iteratively using Newton’s method (argument condition = "catch2"). F as independently estimated parameters is still available with argument condition = "catch".retrospective generic function for SRA objects.nlminb), generating replicates by resampling from the covariance matrix, and filtering non-converged simulation replicates.compare_SRA is a function that compares output and fits from multiple SRA objects with identical model structures in slot SRA@mean_fit but different data weightings, omissions, multipliers, etc.DLMtool::runMSE. Depletion calculations also match those in DLMtool::runMSE.SRA_scope has now been added.SS2OM have been added. The function also generates a markdown report to compare operating model output to Stock Synthesis outputs, e.g., recruitment, catch, spawning biomass time series.SP and SP_SS (surplus production models). This is needed because FMSY is estimated rate parameter rather than r. By default, the minimum CV on the r-prior is 0.1 to allow the model to update r. It is assumed n is fixed in the model.SRA_scope are now more robust (set maximum F in model, higher std. dev. for likelihood of mean lengths).SRA_scope conditioned on either observed catch or observed effort.SRA_scope returns an S4 object of class SRA with a plot() method that generates a markdown report of model fits.SP and SP_SS using life history information (priors in natural mortality and steepness, as well as maturity/weight at age). To use this feature, set argument use_r_prior = TRUE.SP_SS is reduced to 0.1.cDD and cDD_SS are more robust when catch is very, very small (F is set to 0). This is important for management procedures that shut down fishing.make_MP.?MSEtool into the console.multiMSE.SS2OM now has an option for selecting male or female life history parameters.For the new features described below, DLMtool version 5.3.1 is recommended.
multiMSE being the core function. The multiMSE vignette will be quite useful and can be accessed at browseVignettes("MSEtool").Quite a few additions and changes have been made to the Assessment models. See the help manual and vignettes for descriptions of these new Assessment functions.
cDD and cDD_SS, respectively) have been added as new Assessment models to the package. The continuous formulation should be more stable in high F situations.VPA model has also been added to the package.SP assumes continuous production and estimates continuous F’s, similar to ASPIC. This formulation will be more stable in high F situations. The Fox model can be implemented by setting the production function exponent n = 1.spict (state-space surplus production model) has been written and is available in the DLMextra package (located on Github). While reporting functions are available in MSEtool, the output of the wrapper function can still be used with the diagnostic functions in the spict package.SCA and SCA2 estimate annual F’s and include a likelihood function for the catch. In previous versions, SCA matched the predicted catch to observed catch. This feature has been transfered over to the SCA_Pope function.plot function which now generates a markdown report. This will be useful for diagnosing model fits and evaluating parameter estimates.SRA_scope fits an assessment model to catch, indices, and age/length comps to inform historical effort, recruitment deviations, and depletion for data-moderate operating models. Multiple fits are done based on the different life history parameters assumed in the operating model. This function is intended to be an alternative to DLMtool::StochasticSRA.profile and retrospective functions for profiling the likelihood and retrospective analyses, respectively, of assessment models are now improved.compare_models function has been added to compare time series estimates, e.g. B/BMSY and F/FMSY, among different assessment models.start argument) for parameters of assessment models can be expressions and subsequently evaluated in the assessment function. This can be very helpful when passing starting values in the make_MP function.CASAL2OM function can be used to generate an operating model from CASAL assessments.SS2OM and SS2Data functions are updated for the latest versions of r4ss on Github.HCR_ramp).Data-rich-MPHCR_ramp) is now included. Users can input the desired limit and target reference points.make_MP adds dependencies to the MP so that DLMtool::Required returns the appropriate dependencies. Dependencies are dynamic based on the configuration of the assessment model. For example, Data@steep is a dependency for a SCA-based model only if steepness is fixed.simmov function for multiple-area movement models (age-independent)