[ ] A new option to parameterize selectivity in terms of the length-at-50%
[ ] -selection and the slope of the selection ogive has been added because that parameterization seems more stable than the current version, and matches orig better.
[ ] A new retention ogive which forces retention to be zero has been added to GMACS (the old approach in GMACS was to specify a logistic function with a very high length-at-50%-retention). Also, all selectivity functions now have maxima of 1. Orig allowed retention to be less than 1 since 2005. GMACS has been modified to include this option (for example, asymptotic retention is only 0.72 for 2005).
[ ] The prior for NMFS catchability has been updated to match that in orig. In addition, the extra CV associated with the BSFRF survey is now estimated by default (it may have been estimated in some of the sensitivity tests presented previously to the CPT).
[ ] The initial values for the offset of fishing mortality between males and females can be now be specified in the CTL file as well as the phase for the female offsets.
[ ] A new option to specify time-dependent natural mortality as knife-edged changes has been added to better mimic orig. In addition, initial values, bounds and phases for changes in M can be specified in the CTL file. Finally, the ability to mirror natural mortality (i.e. the natural mortality rate parameter for one period can be set to that for another period) has been added, which allows for non-contiguous blocks for natural mortality (an option in orig).
[ ] The years for which recruitment deviations are to be estimated can now be specified in the CTL file.
[ ] Emphasis factors can be specified for the catch by fleet as well as each of the penalties. The values in the CTL are set to best match how orig is currently configured.
[ ] The sample sizes for the length-frequency data were incorrect in the GMACS input files, but have been updated to match those used by orig.
[ ] A new option to parameterize selectivity in terms of the length-at-50%
[ ] -selection and the slope of the selection ogive has been added because that parameterization seems more stable than the current version, and matches orig better.
[ ] A new retention ogive which forces retention to be zero has been added to GMACS (the old approach in GMACS was to specify a logistic function with a very high length-at-50%-retention). Also, all selectivity functions now have maxima of 1. Orig allowed retention to be less than 1 since 2005. GMACS has been modified to include this option (for example, asymptotic retention is only 0.72 for 2005).
[ ] The prior for NMFS catchability has been updated to match that in orig. In addition, the extra CV associated with the BSFRF survey is now estimated by default (it may have been estimated in some of the sensitivity tests presented previously to the CPT).
[ ] The initial values for the offset of fishing mortality between males and females can be now be specified in the CTL file as well as the phase for the female offsets.
[ ] A new option to specify time-dependent natural mortality as knife-edged changes has been added to better mimic orig. In addition, initial values, bounds and phases for changes in M can be specified in the CTL file. Finally, the ability to mirror natural mortality (i.e. the natural mortality rate parameter for one period can be set to that for another period) has been added, which allows for non-contiguous blocks for natural mortality (an option in orig).
[ ] The years for which recruitment deviations are to be estimated can now be specified in the CTL file.
[ ] Emphasis factors can be specified for the catch by fleet as well as each of the penalties. The values in the CTL are set to best match how orig is currently configured.
[ ] The sample sizes for the length-frequency data were incorrect in the GMACS input files, but have been updated to match those used by orig.