build: Format data for model fitting

In pbs-assess/tmbpop: What the Package Does (One Line, Title Case)

Description

Format data for model fitting

Usage

build(survey1, survey2, survey3, paa.catch.female, paa.catch.male,
  n.trips.paa.catch, paa.survey1.female, paa.survey1.male, n.trips.paa.survey1,
  catch, paa.mature, weight.female, weight.male, misc.fixed.param = NULL,
  theta.ini = NULL, lkhd.paa = "normal", var.paa.add = TRUE,
  enable.priors = TRUE)

Arguments

survey1	matrix (or data frame) of biomass estimates from survey 1 (g = 2); each row corresponds to a year of available data, the first column reports the year (T_2), the second column is the survey biomass estimate I t,2 , and the third column is the observation error SD values κ t,2, for t ∈ T_2.
survey2	matrix (or data frame) of biomass estimates from survey 2 (g = 3); each row corresponds to a year of available data, the first column reports the year (T_3), the second column is the survey biomass estimate I t,3 , and the third column is the observation error SD values κ t,3, for t ∈ T_3.
survey3	matrix (or data frame) of biomass estimates from survey 1 (g = 4); each row corresponds to a year of available data, the first column reports the year (T_4), the second column is the survey biomass estimate I t,4 , and the third column is the observation error SD values κ t,4, for t ∈ T_4.
paa.catch.female	matrix (or data frame) of paa for the females (s = 1) from the commercial catch (g = 1); each row corresponds to a year of available data, the first column reports the year (U_1) and the next 30 columns correspond to the paa per age class and year p a,t,1,1 , for t ∈ U_1.
paa.catch.male	matrix (or data frame) of paa for the males (s = 2) from the commercial catch (g = 1); each row corresponds to a year of available data, the first column is the year (U_1) and the next 30 columns correspond to the paa per age class and year p a,t,1,2 , for t ∈ U 1.
n.trips.paa.catch	matrix (or data frame) of the number of trips per year of available paa for the commercial catch (g = 1); each row corresponds to a year, the first column reports the year (U_1), the second column is the number of trips n t,1 , for t ∈ U_1.
paa.survey1.female	matrix (or data frame) of paa for the females (s = 1) from survey 1 (g = 2); each row corresponds to a year of available data, the first column reports the year (U_2) and the next 30 columns correspond to the paa per age class and year p a,t,2,1 , for t ∈ U_2.
paa.survey1.male	matrix (or data frame) of paa for the males (s = 2) from survey 1 (g = 2); each row corresponds to a year of available data, the first column is the year (U_2) and the next 30 columns correspond to the paa per age class and year p a,t,2,2 , for t ∈ U_2.
n.trips.paa.survey1	matrix (or data frame) of the number of trips per year of available paa for survey 1 (g = 2); each row corresponds to a year, the first column reports the year (U_2), the second column is the number of trips n t,2 , for t ∈ U 2.
catch	matrix (or data frame) of commercial catch (g = 1); each row corresponds to a year of available data, the first column reports the year (T_1, i.e the whole range 1940–2012), the second column is the catch C_t.
paa.mature	vector of A = 30 proportions of mature females.
weight.female	vectors of A = 30 average weight at age of females (s = 1).
weight.male	vectors of A = 30 average weight at age of males (s = 2).
misc.fixed.param	optional named vector/list (or data frame) of fixed values for the selectivity parameters of surveys 2 and 3 and for the process error SD σ R ; the names must match muS2, deltaS2, upsilonS2, muS3, deltaS3, upsilonS3 and sigmaR. If unused (left to NULL), then the following default values are used: muS2 = muS3 = 13.3, deltaS2 = deltaS3 = 0.22, upsilonS2 = upsilonS3 = 10 and sigmaR = 0.05.
theta.ini	optional named vector/list (or data frame) of starting values for the model parameters to be estimated; the names must match R0, M1, M2, muC, deltaC, upsilonC, muS1, deltaS1, upsilonS1, h, qS1, qS2 and qS3. If unused (left to NULL), then the default values given in Table 2 are used.
lkhd.paa	either "normal" or "binomial"; specifies a binomial likelihood for the paa, following Cadigan et al. (2014), which may be more statistically sound (and simpler) or a Gaussian likelihood as originally specified in Edwards et al. (2014). Defaults to "normal".
var.paa.add	either TRUE or FALSE; if TRUE (the default), it enables the addition of the 1/(10A) term in Varξ a,t,g,s related to the Gaussian observation error of (F.19 ? ). If FALSE, this additional term is 0 (disabled). If some observed p a,t,g,s are exactly 0 or 1, as in the POP data, then it must be set to TRUE to avoid zero variances with lkhd.paa = "normal".
enable.priors	either TRUE or FALSE; if TRUE (the default), it enables all the prior distributions defined in Section 3.1. If FALSE, the priors are disabled and the maximization over θ is carried without any constraints (apart from constraints on their range of possible values enforced by means of transformations, such as strictly positive variances).

Value

A list of class obj properly formatted to be fed to fit(), with the following elements:

• datalist: a list containing all DATA inputs for the TMB POP template;

• parlist: a list containing all PARAMETER inputs for the TMB POP template;

pbs-assess/tmbpop documentation built on Nov. 5, 2019, 12:16 a.m.