-
Notifications
You must be signed in to change notification settings - Fork 0
/
todo_list.txt
139 lines (91 loc) · 4.94 KB
/
todo_list.txt
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
Tracking down differences in reference points:
THERE IS A BIG DIFFERENCE IN SIZE TRANSITION MATRICES BETWEEN GMACS AND SQ MODELS
many more crab stay inthe same size bin--check the variance on growth increment, adjust/estimate the scale parameter
timing of when MMB is calculated?
Keep using gmacs -phse 6 >a.1 to figure out what's going on with the REF POINTS calcs
Changed how things were mirrored: Must mirror before inside or else inside selectivity gets overwritten.
Problem is that Bzero and B35 are the same for some reason...
CHECK HOW THE LENGTH COMPS ARE BEING CALCULATED...THIS DOESN'T CONSIDER MATURITY YET.
TRY TO AGGREGATE THE MATURE AND IMMATURE--i think that's what's happening right now anyway
FIX the PLOTTING FUNCTION for size comps to deal with immature vs mature
Fix the text saying 50-55% of retained catch are discarded (should be % of total catch are discarded)
PROBLEM: CALCULATING MMB IN A WEIRD WAY. TRY DOING IT NOT BASED ON LOGISTIC, BUT BASED ON A YEARLY PROP MATURE.
How to mimick the way BSFRf?
add 4 fleets: 2009, 2010, NMFS, BSFRF
estimate parameter vector for BSFRF fleet selectivity
mirror NMFS selectivity for NMFS
DOES THAT INCLUDE Q????
Can I mirror Q for indices?
If not, I should put Q in 'get selectivity', instead of 'get index'
Can keep the same Q by giving the same 'index', but a different fleet for selectivity (this is crazy)
MAKE:
a file from survey data that has numbers at length by sex and year to make the ridgeline plots
TEMP FIX WAS TO ALLOW ONLY TRAWL SELX TO NOT BE 1 IN LARGEST BIN
FIX THIS FOR REAL LATER
Discard mortality is applied to numbers, then scaled to biomass in Jack's code.
Something to do with the way discards and mortality/selectivity are calculated.
try to comment out the discard portion of the 'vul' in gmacs?
Trawl selectivity is forcing max to 1.
Fixed this by commenting some code out.
Will need to make an option for selectivity that requires that.
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
Perhaps the retained and total selectivity are reversed from jack's code to gmacs?
Total selectivity is the same-ish.
Retained is not.
Trawl is not.
Jacks code:
all mortality occurs for the year
then growth occurs
then maturity is evaluated based on those
Why woudl this ake a difference for males and not females?
Females are subject to very little F?
Timing of male growth can impact how many are available to the fishery?
This would explain why a bunch are in GMACS, but not in
rec_sdd (size at recruitment) is close, but not exact (to 3 decimal places) and I do not know why.
Is the bloop for males because of when growth occurs relative to the fishery?
Or because of when the n matrices are output?
or is it a selectivity problem?
Big changes--standarad logistic selectivity was logistic95 in most recent version of code
Index is mature first! important for natural mortality
Add in for comparison:
immature/mature natural mortality
change to 4darray
estimate multiplier for immature
add estimated parameter
change ctl file in the
modify calc_natural_mortality
GOTTA FIX CATCH PREDICTIONS
Fix M and Z in project one year ahead and stock recruit and OFL and brute equil
currently they have the maturity index hard coded to 1 for testing
Calc predicted catch (+ _out) too--search for "Z(" to find instance and make sure they are not hard coded
Add in for fitting:
different survey q by sex
Set a .PIN file based on not_gmacs
Run one function call
Compare the N matrices
Check timing of different processes between models
fix calc size comps function so they can be split out by maturity/shell
tALK TO PEOPLE ABOUT FEMALE DISCARDS--selectivity is so much different and hard to figure out 'total' selectivity and 'female' selectivity
I don't understand the mechanism here.
it seems like this suggests that selectivity means 'availability' here...and there is targeting
Check to see the timing of molting maturity between the two models:
is the probability of maturity from 2016sc based on the postmolt or premolt?
shell_force: does not have terminal molt, uses shell condition to get in mature/immmature
comp_n: trying to match output in not_gmacs_2, uses code from term_molt
not_gmacs_2: current assessment code, but using simplified assumptions (single recdevs, etc.)
term_molt: introduces terminal molt to GMACS
Errors to build in:
nmature and nshell cannot both be 2
processes:
recruitment (sex ratio equal, calculate initial recruitment from estimated initial N)
recruit at size (output the proportions per bin, estimate parameters based on gmacs code)
selectivity (get parameters in the same space)
growth (input growth increments; might have to calculate the
fishing mortality (where do I input this?)
molting probability (input)
natural mortality
Start thinking about steps to get assessment running:
managment output
how to incorporate growth (skip to linear after you show GMACS can reproduce dynamics)
how to incorporate BSFRF data
how to incorporate maturity data