turn off extra switches when Do_More_Std == 1

[iantaylor-NOAA]

Concisely describe what has been changed/addressed in the pull request.

@e-perl-NOAA figured out that the Mac runner for r4ss GitHub actions were failing while Windows and Linux were passing. She was able to export the echoinput which ended with do SmryBio std labels. However, I noticed that this line should only occur if the # extra stdev reporting was 2 but in this model it was 1. Digging into the SS3 source code revealed that some switches were never set for the case where Do_More_Std == 1, and apparently the Mac filled in the missing value differently than the other operating systems.

What tests have been done?

Where are the relevant files?

@e-perl-NOAA checked this build from this branch to confirm that it resolved the issue.

What tests/review still need to be done?

None.

Is there an input change for users to Stock Synthesis?

[x] No, there was no input change.

[Rick-Methot-NOAA]

alternatively:
remove the line 5589 so everything gets null value before getting to line 5608.

[e-perl-NOAA]

We may have to try that since the r4ss model retrospectives stopped here in the echoinput file with Ian's changes

 Biology base parameter setup 22
....
....
21  0.1 10 1 1 1 0 -1 0 0 0 0 0 0 0 CohortGrowDev
22  1e-06 0.999999 0.5 0.5 0.5 0 -99 0 0 0 0 0 0 0 FracFemale_GP_1

[Rick-Methot-NOAA]

the code to do retrospectives is not smart enough to know to turn off request for se of output that no longer exists when a year disappears due to the retrospective peel. Adding that extra intelligence to the code might be more than trivial.

[iantaylor-NOAA]

@Rick-Methot-NOAA, thanks for looking this over. I just pushed the change to line 5589 as you suggested.
But it sounds like this won't necessarily fix the retro problem.

[Rick-Methot-NOAA]

I got i fixed but in looking at code I see that some quantities, like natage_std_yr are reported for endyr. This is ok even in retro because SS3 goes through all years in retro. But we could consider whether to provide that se output for retro_yr, rather than endyr

[e-perl-NOAA]

I will report back with the r4ss testing on the retro with simple_small when that test finishes running.

[e-perl-NOAA]

echoinput.txt
Sadly, it's still stopping after the MG parameters are read for the mac.

[Rick-Methot-NOAA]

echo should be showing mgparm_type for each parm at that stage of the file read. I'll investigate as it could be another unitialized variable probably in this section of code: if (recr_dist_method < 4) mgp_type(Ip, MGP_CGD - 1) = 4; // recruit apportionments mgp_type(MGP_CGD) = 2; // cohort growth dev if (do_migration > 0) mgp_type(MGP_CGD + 1, N_MGparm) = 5; // note that it fills until end of MGparm list, but some get overwritten if (N_platoon > 1 && sd_ratio_rd < 0) mgp_type(sd_ratio_param_ptr) = 2; if (Use_AgeKeyZero > 0) mgp_type(AgeKeyParm, N_MGparm) = 6; if (catch_mult_pointer > 0) mgp_type(catch_mult_pointer, N_MGparm) = 7; for (f = frac_female_pointer; f <= frac_female_pointer + N_GP - 1; f++) mgp_type(f) = 4; if (N_pred > 0) mgp_type(predparm_pointer(1), predparm_pointer(1) + N_predparms - 1) = 1; echoinput << "mgparm_type for each parm:"<<endl;

[Rick-Methot-NOAA]

fixed issue