Unify F approach

[Rick-Methot-NOAA]

continues development for issue #570 and especially #581
See description and example in #581
Substantial refactoring of the internal machinery for managing F as parameter vs F as hybrid
extends the F-detail approach previously available only for F_method 2 to F_method 4

only new input is for F_method 4 if using the F_detail feature.

[e-perl-NOAA]

Okay so the spinydogfish model echoinput file is stopping after:

GP Seas Area Use? female_recr_dist
1 1 1 1 0.5
 made the age_age' key 
Finished evaluating biology calculations once


Begin estimating

Warning file looks like the following:

Note 1 Information: Max data length bin: 132  < max pop len bins: 136; so will accumulate larger pop len bins
3 iter 1
 hrate for 1970: 

[Rick-Methot-NOAA]

that stop in dogfish might only be because I had not yet removed a specific output statement while debugging. It is gone now.

[e-perl-NOAA]

Yes, all seems to be working!

[iantaylor-NOAA]

Looks good. I didn't try to follow all the details, but I don't see any red flags. The tests are all passing, including the vermillion_snapper_F4 model which uses F Method 4.

I compared Report files and see the DISCARD_AT_AGE is now available in the forecast as intended (discussed in #569).

I see control.ss_new remains the same except for improved comments

[iantaylor-NOAA]

For a long-lived stock, would setting a smaller max harvest rate ever be a useful tool for constraining the model to a reasonable parameter space, or is the ballpark F a better approach and max F should never be constraining?

[Rick-Methot-NOAA]

Thanks Ian. Nothing magic about the 2.9 value. The role of max F is mostly to keep model from exploring F levels that nearly extirpate older fish. In the attached example, it is constraining during early hybrid phases to keep F below 2.9, which creates a catch penalty, which moves model back to higher biomass and lower F values. For long-lived stocks a maxF of 0.4 might seem reasonable, but even 0.40 will not extirpate.