search

PIFSCstockassessments / ss3diags

R package with advanced diagnostics to evaluate a Stock Synthesis model. Diagnostics include residual analyses, hindcasting and cross-validation techniques, and retrospective analyses.
http://pifscstockassessments.github.io/ss3diags/
2 stars 4 forks source link

SSplots: "summaryoutput" variable inconsistency for "timeseries" SSplots #70

Open efletcherPIFSC opened 1 year ago

efletcherPIFSC commented 1 year ago

SSplotEnsemble, SSplotModelComp, SSplotHCXval, and SSplotRetro have the primary (or first) argument, with different names, and documented differently. However these values are very simlar: they are derived from r4ss::SSsummarize .

All four functions would have the primary argument as summaryoutput defined as "List created by r4ss::SSsummarize()". SSplotEnsemble and SSplotHCXval will need to rename their variables to summaryoutput, but it is the same functionality. However, does these changes affect the context of any of the function documentation?

MOshima-PIFSC commented 1 year ago

Hi Eric, I don't think the SSplotEnsemble() kb object is the same as the summaryoutput object. The kb object is created from the SSdeltaMVLN() function and has a different structure than the other summaryoutput objects. I think it is ok to change retroSummary to summaryoutput but it might be helpful to add a note somewhere that it is the summarized list from the retrospective runs and if they are plotting age or length composition data, then they will need to use ss3diags::SSretroComps() instead of r4ss::SSsummarize(). This note should probably be added to SSplotRetro() as well.