Add data_small dataset and associated test and documentation

.github/workflows/build-doxygen.yml

@@ -21,7 +21,7 @@ jobs:
 
     steps:
     - name: Get repository
-      uses: actions/checkout@v4
+      uses: actions/checkout@v5
 
     - name: Installing build dependencies
       run: |

.github/workflows/call-allcontributors.yml

@@ -19,7 +19,7 @@ jobs:
             pull-requests: write
         steps:
             - name: Checkout repository
-              uses: actions/checkout@v4
+              uses: actions/checkout@v5
             - name: Setup R
               uses: r-lib/actions/setup-r@v2
 

.github/workflows/get-gtest-codecov.yml

@@ -19,7 +19,7 @@ jobs:
       GITHUB_PAT: ${{ secrets.GITHUB_TOKEN }}
 
     steps:
-    - uses: actions/checkout@v4
+    - uses: actions/checkout@v5
     - name: Install Ninja
       run: sudo apt-get install ninja-build
     - uses: r-lib/actions/setup-r@v2

.github/workflows/greetings.yml

@@ -2,14 +2,18 @@ name: Greetings
 
 on: [pull_request_target, issues]
 
+permissions:
+  issues: write
+  pull-requests: write
+
 jobs:
   greeting:
     runs-on: ubuntu-latest
     permissions:
       issues: write
       pull-requests: write
     steps:
-    - uses: actions/first-interaction@v1
+    - uses: actions/first-interaction@v3
       with:
         repo-token: ${{ secrets.GITHUB_TOKEN }}
         issue-message: >

.github/workflows/run-clang-format.yml

@@ -27,7 +27,7 @@ jobs:
     # Format hpp and cpp files under inst/include, src/, and test/gtest
     # We use Google style to format code.
     steps:
-    - uses: actions/checkout@v4
+    - uses: actions/checkout@v5
     - uses: DoozyX/clang-format-lint-action@v0.20
       with:
         source: './inst/include ./src ./tests/gtest'

.github/workflows/run-clang-tidy.yml

@@ -27,7 +27,7 @@ jobs:
     strategy:
       fail-fast: false
     steps:
-    - uses: actions/checkout@v4
+    - uses: actions/checkout@v5
 
     - name: install clang-tidy
       run: sudo apt update && sudo apt -y install clang-tidy

.github/workflows/run-googletest.yml

@@ -16,7 +16,7 @@ jobs:
       GITHUB_PAT: ${{ secrets.GITHUB_TOKEN }}
 
     steps:
-    - uses: actions/checkout@v4
+    - uses: actions/checkout@v5
     - name: Install Ninja
       shell: bash
       run: ${{ runner.os == 'macOS' && 'brew install ninja' || runner.os == 'Windows' && 'choco install ninja' || 'sudo apt-get install ninja-build' }}

CMakeLists.txt

@@ -54,33 +54,6 @@ IF(BUILD_DOC)
     COMMENT "Building documentation with Doxygen.")
 ENDIF()
 
-# Set up Rcpp
-FetchContent_Declare(
-  rcpp
-  URL https://github.com/RcppCore/Rcpp/archive/refs/tags/1.0.8.zip
-)
-FetchContent_MakeAvailable(rcpp)
-
-# Set up R
-MESSAGE(STATUS "Looking for R executable")
-
-EXECUTE_PROCESS(
-  COMMAND which R
-  OUTPUT_VARIABLE R_EXECUTABLE)
-
-MESSAGE(STATUS "R_EXECUTABLE is ${R_EXECUTABLE}")
-
-EXECUTE_PROCESS(
-		COMMAND R "--slave" "--no-save" "-e" "cat(R.home())"
-		OUTPUT_VARIABLE R_HOME)
-
-MESSAGE(STATUS "R_HOME is ${R_HOME}")
-
-EXECUTE_PROCESS(
-		COMMAND R "--slave" "--no-save" "-e" "remotes::install_local(upgrade = 'always')")
-EXECUTE_PROCESS(
-		COMMAND R "--slave" "--no-save" "-e" "FIMS:::setup_gtest()")
-
 # Enable testing for current directory and below
 enable_testing()
 
@@ -91,8 +64,6 @@ add_library(fims_test INTERFACE)
 target_include_directories(fims_test
   INTERFACE
     inst/include
-    ${rcpp_SOURCE_DIR}/inst/include
-    ${R_HOME}/include
 )
 
 # Add compile definition STD_LIB to the fims_test target.

NAMESPACE

@@ -64,7 +64,6 @@ export(initialize_process_distribution)
 export(initialize_process_structure)
 export(inv_logit)
 export(is.FIMSFit)
-export(is.FIMSFits)
 export(log_error)
 export(log_info)
 export(log_warning)

R/check_fims.R

@@ -66,8 +66,7 @@ setup_gtest <- function() {
 #' @keywords developer
 #' @examples
 #' \dontrun{
-#' # TODO: add examples of strings passed to `...`
-#' setup_and_run_gtest()
+#' setup_and_run_gtest("--rerun-failed --output-on-failure")
 #' }
 setup_and_run_gtest <- function(...) {
   setup_gtest()

R/data.R

@@ -0,0 +1,76 @@
+#' FIMS input data frame
+#'
+#' A dataset containing information necessary to run an age-structured stock
+#' assessment model in FIMS. This data was generated using
+#' the `ASSAMC` package written for the [model comparison project](
+#' www.github.com/NOAA-FIMS/Age_Structured_Stock_Assessment_Model_Comparison).
+#'
+#' @format
+#' A data frame with `r NROW(data1)` observations of `r NCOL(data1)`
+#' variables:
+#' \describe{ \item{type}{The type of data the row contains. Allowed types
+#'   include `age`, `length`, `index`, `landings`, `age-to-length-conversion`,
+#'     and `weight-at-age` data.}
+#'   \item{name}{A character string providing the name of the information source
+#'     that the data was collected from, e.g., `"Trawl fishery"`.}
+#'   \item{age}{An integer age. Entry can be `NA` if information pertains to
+#'     multiple ages, e.g., total landings rather than landings of age-4 fish.}
+#'   \item{length}{A numeric length. Entry can be `NA` if information doesn't
+#'     pertain to length.}
+#'   \item{datestart,dateend}{Start and end dates of the data collection period.
+#'     Format all dates using `yyyy-mm-dd`, which can accommodate fake years
+#'     such as `0001-01-01`.}
+#'   \item{value}{The measurement of interest.}
+#'   \item{unit}{A character string specifying the units of `value`. Allowed
+#'     units for each data type are as follows. `mt` is used for `index`,
+#'     `landings`, and `weight-at-age` data. `number` or `proportion` are each
+#'     viable units for the composition data, where the former is the preferred
+#'     unit of measurement.}
+#'   \item{uncertainty}{A real value providing a measurement of uncertainty
+#'     for value. For landings and indices of abundance this should be
+#'     the standard deviation of the logged observations if you are using the
+#'     lognormal distribution to fit your data. For composition data it will
+#'     be your input sample size.
+#' }
+#' }
+#' @source \url{www.github.com/NOAA-FIMS/Age_Structured_Stock_Assessment_Model_Comparison}
+"data1"
+
+#' small FIMS input data frame
+#'
+#' A dataset containing minimal information necessary to run an 
+#' age-structured stock
+#' assessment model in FIMS. This data was generated using the script in 
+#' `/data-raw/data_small.R` and does not correspond to any realistic population
+#' or sampling method (the values are mostly arbitrary).
+#'
+#' @format
+#' A data frame with `r NROW(data_small)` observations of `r NCOL(data_small)`
+#' variables:
+#' \describe{ \item{type}{The type of data the row contains. Allowed types
+#'   include `age`, `length`, `index`, `landings`, `age-to-length-conversion`,
+#'     and `weight-at-age` data.}
+#'   \item{name}{A character string providing the name of the information source
+#'     that the data was collected from, e.g., `"Trawl fishery"`.}
+#'   \item{age}{An integer age. Entry can be `NA` if information pertains to
+#'     multiple ages, e.g., total landings rather than landings of age-4 fish.}
+#'   \item{length}{A numeric length. Entry can be `NA` if information doesn't
+#'     pertain to length.}
+#'   \item{datestart,dateend}{Start and end dates of the data collection period.
+#'     Format all dates using `yyyy-mm-dd`, which can accommodate fake years
+#'     such as `0001-01-01`.}
+#'   \item{value}{The measurement of interest.}
+#'   \item{unit}{A character string specifying the units of `value`. Allowed
+#'     units for each data type are as follows. `mt` is used for `index`,
+#'     `landings`, and `weight-at-age` data. `number` or `proportion` are each
+#'     viable units for the composition data, where the former is the preferred
+#'     unit of measurement.}
+#'   \item{uncertainty}{A real value providing a measurement of uncertainty
+#'     for value. For landings and indices of abundance this should be
+#'     the standard deviation of the logged observations if you are using the
+#'     lognormal distribution to fit your data. For composition data it will
+#'     be your input sample size.
+#' }
+#' }
+#' @source None, this is a simulated dataset full of arbitrary values as defined in `/data-raw/data_small.R`.
+"data_small"

R/distribution_formulas.R

@@ -290,7 +290,7 @@ initialize_data_distribution <- function(
     new_module <- methods::new(DlnormDistribution)
 
     # populate logged standard deviation parameter with log of input
-    # Using resize() and then assigning value to each element of log_sd diretly
+    # Using resize() and then assigning value to each element of log_sd directly
     # is correct, as creating a new ParameterVector for log_sd here would
     # trigger an error in integration tests with wrappers.
     new_module$log_sd$resize(length(sd[["value"]]))

R/fimsfit.R

@@ -22,11 +22,6 @@ utils::globalVariables(c(
 # TODO: Fix "no metadata object found to revise superClass" in sdreportOrList
 # TODO: Write more validity checks for FIMSFit
 # TODO: Better document the return of [get_estimates()], i.e., columns
-# TODO: Decide if the error from is.FIMSFits should be a single FALSE or stop
-# TODO: Decide if "total" should be a part of number_of_parameters because it
-#       can be calculated from fixed_effects + random_effects and would need to
-#       be calculated in print.FITFims()
-# TODO: Determine if report should always use last.par.best
 # TODO: Make a helper function to add lower and upper CI for users in estimates
 
 # methods::setClass: ----
@@ -95,6 +90,12 @@ methods::setMethod(
       x@number_of_parameters,
       sep = "="
     )
+    total_parameters <- sum(x@number_of_parameters)
+    all_parameters_info <- c(number_of_parameters, paste(
+      "total",
+      total_parameters,
+      sep = "="
+    ))
     div_digit <- cli::cli_div(theme = list(.val = list(digits = 5)))
     terminal_ssb <- sapply(
       x@report[["ssb"]],
@@ -103,7 +104,7 @@ methods::setMethod(
     cli::cli_inform(c(
       "i" = "FIMS model version: {.val {x@version}}",
       "i" = "Total run time was {.val {rt}} {ru}",
-      "i" = "Number of parameters: {number_of_parameters}",
+      "i" = "Number of parameters: {all_parameters_info}",
       "i" = "Maximum gradient= {.val {x@max_gradient}}",
       "i" = "Negative log likelihood (NLL):",
       "*" = "Marginal NLL= {.val {x@opt$objective}}",
@@ -343,21 +344,6 @@ is.FIMSFit <- function(x) {
   inherits(x, "FIMSFit")
 }
 
-#' Check if an object is a list of FIMSFit objects
-#'
-#' @param x List of fits returned from multiple calls to [fit_fims()].
-#' @keywords fit_fims
-#' @export
-is.FIMSFits <- function(x) {
-  if (!is.list(x)) {
-    cli::cli_warn(
-      message = c("x" = "{.par x} is not a list -- something went wrong.")
-    )
-    return(FALSE)
-  }
-  all(sapply(x, function(i) inherits(i, "FIMSFit")))
-}
-
 # Constructors ----
 
 #' Class constructors for class `FIMSFit` and associated child classes
@@ -431,7 +417,6 @@ FIMSFit <- function(
   n_fixed_effects <- length(obj[["par"]])
   n_random_effects <- length(obj[["env"]][["parList()"]][["re"]])
   number_of_parameters <- c(
-    total = n_total,
     fixed_effects = n_fixed_effects,
     random_effects = n_random_effects
   )
@@ -712,6 +697,13 @@ fit_fims <- function(input,
                        trace = 0
                      ),
                      filename = NULL) {
+  # See issue 455 of sdmTMB to see what should be used.
+  # https://github.com/pbs-assess/sdmTMB/issues/455
+  # NOTE: When we add implementation for newton step we need to
+  # review the above github issue to make sure we maintain continuity
+  # between outputs as last.par may not equal last.par.best due to
+  # the smallest newton gradient solution not matching the smallest
+  # likelihood value. This can cause sanity issues in output reporting.
   if (number_of_newton_steps > 0) {
     cli::cli_abort("Newton steps not implemented yet.")
   }

R/initialize_modules.R

@@ -820,15 +820,6 @@ initialize_fims <- function(parameters, data) {
         data_type = "landings"
       )
     }
-    # TODO (Matthew): Determine if the "dims" field is required for DmultinomDistribution.
-    # We need to decide whether to:
-    # 1. Remove the "dims" field to maintain consistency with other distributions, or
-    # 2. Update all relevant R functions (e.g., initialize_data_distribution())
-    #    that call DmultinomDistribution to set the "dims" field.
-    # AMH comment:
-    # 1. dims field is needed to track the dimension of multivariate input
-    # 2. the pattern for multinomial is different because it is a multivariate distribution.
-    #  Other multivariate distributions (e.g. MVNORM) will likely also need a dims field
 
     if ("age" %in% fleet_types &&
       "AgeComp" %in% data_distribution_names_for_fleet_i) {

R/reshape_output.R

@@ -120,7 +120,7 @@ reshape_tmb_estimates <- function(obj,
                                   opt = NULL,
                                   parameter_names) {
   # Outline for the estimates table
-  # TODO: The fleet_name, age, length, and time columns are currently emplty. Matthew
+  # TODO: The fleet_name, age, length, and time columns are currently empty. Matthew
   # has started adding information to the JSON output in the dev-model-families branch.
   # We can populate these columns once the dev-model-families branch is merged
   # into dev.