[Feature]: Develop design matrix approach to share movement parameters

[k-doering-NOAA]

Imported from redmine, Issue #70982
Opened by @RickMethot on 2019-11-06
Status when imported: Under Review

males and females can show different seasonal movement patterns. There needs to be a way to specify this

[k-doering-NOAA]

comment from @RickMethot on 2019-11-07:
better to do this to allow movement to be specific to sex, GP, settlement
best to do this as a list of movement patterns and to refer to this from a table by sex, gp, settlement
so like the tag parameter approach

[k-doering-NOAA]

comment from @RickMethot on 2019-11-08:
richard.methot wrote:

better to do this to allow movement to be specific to sex, GP, settlement
best to do this as a list of movement patterns and to refer to this from a table by sex, gp, settlement
so like the tag parameter approach

1. modify do_migration such that -999 triggers reading in alternative format
2. read list of migration definitions. Each specifies: season, GP, sex, settlement, source, destination, functional_form, parameter_set
   (a) where functional_form will allow for future elaboration of forms other than the current two-parm dogleg
   (b) terminate list with -999 for season
   (c) if parameter_set is shared by two of more definitions, they must specify same functional_form
   (d) find the max parameter_set requested
   (e) issue warning if a parameter_set ID is skipped
   (f) assign to a morph ID that refers to sexGPsettlement: GP3(g)
   (g) then for each g create pointer to which movement pattern it will

IDEA: make way for sharing of young fish parameter but have unique old fish parameter, or vice versa

3. read and name parameters according to the parameter_sets and functional_forms

[k-doering-NOAA]

Yukio Takeuchi has the following request:

In the movement section (78-89pp) of SS 3.30.16 user manual, there are two
special values -9999 and -9998. The former is mention to be used to the case
where same movement rate as defined the first defined movement pattern. The
latter is mentioned to be used with the case when same movement rate is to
be applied for age1 and age2 of same movement pattern. The latter works
pretty well for my model and contributed to reduce the number of parameters.

I am wondering if I can share(mirror) movement rate parameter of different
growth pattern (e.g. I am currently considering to assign 1 growth for each
season. I.e., settlements occurring season 1 correspond to GP1 and
settlements occurring season 2 correspond to GP2 and so on. In this case is
it possible to set up mirroring of movement rate parameters for movement
pattern of different GP but sharing same source and destination?  As long as
I read the user manual and source code related to the movement rate
(FUNCTION void get=migration of ss=3.30.16.tpl), I can not figure out how to
mirror movement rate across growth pattern.

Just changed the title of this issue to reflect that the approach being developed would not only be to share movement params across sexes, but could also be across other partitions, such as growth patterns.