Model fit (#358)

* started new model design function

* started fishset model design function

* started format data function

* working on formal data function, but need to figure out create_distance_matrix

* working on updating create_dist_matrix

* cleaned up code in create_distance_matrix but still need to run tests and make output consistent

* refactored create_distance_matrix

* added unit tests for create_dist_matrix and updated documentaiton

* updated model design section for creating distance matrix

* added 'long format data' to data types for unserialize_table

* format_model_deisgn completed but need some minor changes

* finished format_model_data() and added unit tests

* finished fishset_design function

* added unit tests for fishset_design

* clean up comments for testing

* fixed issues with unit tests

* commented out modeling code in vignette for now

* mostly completed the fishset_fit() function

* added sql table save to fishset_fit

* added unit tests for fishset_fit

* added RTMB to DESCRIPTION file

* remove NULL default for alt_name

* fixed mismatch in data class types

* added code to check class of unique ID

* added a 'native' print function for the fishset_fit function

* addressing PR review comments

* fix unit tests for format_model_data

* fix merge conflict in globals.r

* clean up more conflicts

* added error check for unique obs id in format_model_data

* minor updates

* updated unit tests

* removed a bunch of old modeling functions

* delete a bunch of old modeling functions

* addressed PR comments, cleaned up some R CMD warnings

Author: Paul-Carvalho

DESCRIPTION

@@ -72,6 +72,7 @@ Imports:
     readxl,
     rlang,
     RSQLite,
+    RTMB,
     scales,
     servr,
     sf (>= 1.0-0),

NAMESPACE

@@ -1,5 +1,6 @@
 # Generated by roxygen2: do not edit by hand
 
+S3method(print,fishset_fit)
 export(ID_var)
 export(aboutUI)
 export(add_prompter)
@@ -36,13 +37,10 @@ export(create_dist_between_for_gui)
 export(create_dist_matrix)
 export(create_duration)
 export(create_expectations)
-export(create_logit_input)
 export(create_mid_haul)
-export(create_model_input)
 export(create_proj_settings)
 export(create_seasonal_ID)
 export(create_var_num)
-export(cross_validation)
 export(current_db_table)
 export(current_log)
 export(current_out)
@@ -63,7 +61,6 @@ export(delete_meta)
 export(delete_models)
 export(density_plot)
 export(deparse_name)
-export(discretefish_subroutine)
 export(dms_to_dd)
 export(dms_to_pdms)
 export(dummy_matrix)
@@ -75,16 +72,12 @@ export(empty_vars)
 export(empty_vars_filter)
 export(epm_lognormal)
 export(epm_normal)
-export(epm_predict)
 export(epm_weibull)
 export(erase_project)
 export(exp_catch_names)
 export(expand_data)
 export(expected_catch_list)
-export(explore_startparams)
-export(explore_startparams_discrete)
 export(filter_dat)
-export(filter_outsample)
 export(filter_summary)
 export(filter_table)
 export(find_catch)
@@ -104,6 +97,7 @@ export(find_project)
 export(find_value)
 export(fishset_compare)
 export(fishset_design)
+export(fishset_fit)
 export(fishset_tables)
 export(fishset_theme)
 export(fishset_viridis)
@@ -112,7 +106,6 @@ export(fleet_assign)
 export(fleet_table)
 export(format_grid)
 export(format_model_data)
-export(format_outsample_coefs)
 export(freq_table)
 export(function_summary)
 export(get_closure_scenario)
@@ -161,13 +154,9 @@ export(log_func_model)
 export(log_rerun)
 export(log_rerun_gui)
 export(log_reset)
-export(logit_c)
 export(logit_correction)
-export(logit_predict)
-export(logit_zonal)
 export(lonlat_to_centroid)
 export(main_tables)
-export(make_model_design)
 export(make_spat_area)
 export(map_kernel)
 export(map_plot)
@@ -196,9 +185,7 @@ export(meta_list)
 export(meta_log_call)
 export(meta_tables)
 export(metadata_gui)
-export(mixed_logit_predict)
 export(model_design_list)
-export(model_design_outsample)
 export(model_error_summary)
 export(model_fit)
 export(model_fit_summary)
@@ -239,7 +226,6 @@ export(policy_metrics)
 export(polyval)
 export(pred_prob_outputs)
 export(predict_map)
-export(predict_outsample)
 export(predict_probability)
 export(pretty_lab)
 export(pretty_tab)
@@ -351,7 +337,6 @@ importFrom(DBI,dbExistsTable)
 importFrom(DBI,dbGetQuery)
 importFrom(DBI,dbListFields)
 importFrom(DBI,dbListTables)
-importFrom(DBI,dbReadTable)
 importFrom(DBI,dbRemoveTable)
 importFrom(DBI,dbWriteTable)
 importFrom(DT,DTOutput)
@@ -365,6 +350,10 @@ importFrom(Formula,Formula)
 importFrom(Formula,model.part)
 importFrom(RCurl,getURL)
 importFrom(RSQLite,SQLite)
+importFrom(RTMB,MakeADFun)
+importFrom(RTMB,REPORT)
+importFrom(RTMB,getAll)
+importFrom(RTMB,sdreport)
 importFrom(Rcpp,sourceCpp)
 importFrom(data.table,":=")
 importFrom(data.table,.GRP)
@@ -564,6 +553,7 @@ importFrom(stats,as.formula)
 importFrom(stats,ave)
 importFrom(stats,coef)
 importFrom(stats,complete.cases)
+importFrom(stats,cov2cor)
 importFrom(stats,dbinom)
 importFrom(stats,density)
 importFrom(stats,dexp)
@@ -578,7 +568,8 @@ importFrom(stats,median)
 importFrom(stats,model.matrix)
 importFrom(stats,na.omit)
 importFrom(stats,na.pass)
-importFrom(stats,optim)
+importFrom(stats,nlminb)
+importFrom(stats,pchisq)
 importFrom(stats,pnorm)
 importFrom(stats,predict)
 importFrom(stats,quantile)

R/check_exp.R

@@ -2,12 +2,12 @@ check_exp <- function(ec, ec_names) {
 #' Prepare expected catch matrix
 #' 
 #' Checks that specified expected catch matrices exists and is formatted correctly
-#' when \code{\link{make_model_design}} is called.
+#' when \code{make_model_design} is called.
 #' 
 #' @param ec Expected catch list created by \code{\link{create_expectations}}.
 #' @param ec_names The names of the expected catch matrices to include in the
 #'   model and how they are used (i.e. used together or separately). See the
-#'   \code{expectcatchmodels} argument in \code{\link{make_model_design}}.
+#'   \code{expectcatchmodels} argument in \code{make_model_design}.
 #' @returns Returns a list containing the filtered expected catch list and specified
 #'   matrices to include in model.
 #' @details Checks that \code{ec_names} is specified properly by identifying 

R/create_alternative_choice.R

@@ -81,25 +81,25 @@
 #' @details Defines the alternative fishing choices. These choices are used to develop
 #'  the matrix of distances between observed and alternative fishing choices (where
 #'  they could have fished but did not). The distance matrix is calculated by the
-#'  [make_model_design()] function. `occasion` defines the observed fishing
+#'  [format_model_data()] function. `occasion` defines the observed fishing
 #'  location and `alt_var` the alternative fishing location. `occasion_var`
 #'  identifies an ID column or set of lon-lat variables needed to create the
 #'  distance matrix.
 #'
 #'  Parts of the alternative choice list are pulled by [create_expectations()],
-#'  [make_model_design()], and the model run [discretefish_subroutine()])
+#'  [format_model_data()], and the model run [fishset_fit()])
 #'  functions. These output include choices of which variable to use for catch and
 #'  which zones to include in analyses based on a minimum number of hauls per trip
 #'  within a zone. Note that if the alternative choice list is modified, the
-#'  [create_expectations()] and [make_model_design()] functions
+#'  [create_expectations()] and [format_model_data()] functions
 #'  should also be updated before rerunning models.
 #' 
 #' @return Function saves a list of alternative choice matrices to the FishSET
 #'  database as `projectAlternativeChoice`. The list includes
 #'  the alternative choice list from the user-defined choices. Multiple alternative
 #'  choice cases can be added to the list by specifying unique names. The list is
 #'  automatically saved to the FishSET database and is called in
-#'  `make_model_design`.
+#'  `format_model_data`.
 
 create_alternative_choice <- function(dat,
                                       project,

R/create_dist_matrix.R

@@ -28,7 +28,7 @@
 #' @importFrom stats setNames
 #' @export 
 #' @keywords internal
-#' @details Function is called by \code{\link{make_model_design}} to generate the
+#' @details Function is called by \code{\link{format_model_data}} to generate the
 #'   distance matrix. Alternative fishing options come from the Alternative Choice
 #'   list, generated from the \code{\link{create_alternative_choice}} function.
 #' @return
@@ -251,7 +251,7 @@ create_dist_matrix <-
     if (is.null(o_var)) o_var <- occasion_var
 
     return(list(dist_matrix = dist_matrix, 
-                alt_choice_units = units, 
+                dist_units = units, 
                 alt_choice_type = "distance", 
                 occasion = occasion,
                 occasion_var = o_var,

R/create_expectations.R

@@ -1,7 +1,7 @@
 #' Create expected catch/expected revenue matrix
 #'
 #' Create expected catch or expected revenue matrix. The matrix is required for 
-#' the \code{\link{logit_c}} model. Multiple matrices (with unique names) can be saved in a 
+#' the conditional logit model. Multiple matrices (with unique names) can be saved in a 
 #' project.
 #' 
 #' @param dat Primary data containing information on hauls or trips. Table in FishSET 
@@ -55,11 +55,11 @@
 #'   as \code{projectExpectedCatch}. The list includes 
 #'   the expected catch matrix from the user-defined choices. Multiple expected catch cases 
 #'   can be added to the list by specifying unique names. The list is automatically saved to 
-#'   the FishSET database and is called in \code{\link{make_model_design}}. The expected catch 
+#'   the FishSET database and is called in \code{\link{format_model_data}}. The expected catch 
 #'   output does not need to be loaded when defining or running the model.
 #' @details Function creates an expectation of catch or revenue for alternative 
 #'   fishing zones (zones where they could have fished but did not). The output is 
-#'   saved to the FishSET database and called by the \code{\link{make_model_design}} 
+#'   saved to the FishSET database and called by the \code{\link{format_model_data}} 
 #'   function. \code{\link{create_alternative_choice}} must be called first as observed 
 #'   catch and zone inclusion requirements are defined there.\cr 
 #'   The primary choices are whether to treat data as a fleet or to group the data