fix(data_mile1): sd of log(x) instead of CV

For catch and index data the measurement of uncertainty should be the
standard deviation (sd) of the logged observations. The operating model
outputs a CV. I added a temporary function in data_mile1.R to calculate
the sd from the CV. In the future we will need to, behind the scenes,
take the log of the sd so the parameters stay within the bounds, which
is what the vignette is doing now. This time series of uncertainty can
later be fed to the modules but I did not change the vignette.

Close #136

R/data_mile1.R

@@ -24,7 +24,11 @@
 #'     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 value. For catches and survey 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/Bai-Li-NOAA/Age_Structured_Stock_Assessment_Model_Comparison}
 "data_mile1"

data-raw/data_mile1.R

@@ -15,6 +15,10 @@
 ###############################################################################
 # Helper functions and load packages
 ###############################################################################
+cv_2_sd <- function(x) {
+  sqrt(log(x^2 + 1))
+}
+
 check_ASSAMC <- function() {
   packages_all <- .packages(all.available = TRUE)
   if (!"ASSAMC" %in% packages_all) {
@@ -25,6 +29,7 @@ check_ASSAMC <- function() {
   library("ASSAMC")
   return(TRUE)
 }
+
 check_ASSAMC()
 library(dplyr)
 
@@ -52,11 +57,7 @@ landings_data <- data.frame(
   ),
   value = returnedom[["em_input"]]$L.obs[[1]],
   unit = "mt", # metric tons
-  # TODO: discuss if CV the appropriate input here given that landings will be
-  #       modeled with a lognormal or similar likelihood. Just because previous
-  #       models have used CV doesn't mean we have to continue to use it.
-  #       E.g., `dlnorm(sdlog = )` uses a standard deviation on the log scale.
-  uncertainty = returnedom[["em_input"]]$cv.L[[1]]
+  uncertainty = cv_2_sd(returnedom[["em_input"]]$cv.L[[1]])
 )
 
 ###############################################################################
@@ -76,7 +77,7 @@ index_data <- data.frame(
   ),
   value = returnedom[["em_input"]]$surveyB.obs[[1]],
   unit = "mt",
-  uncertainty = returnedom[["em_input"]]$cv.survey[[1]]
+  uncertainty = cv_2_sd(returnedom[["em_input"]]$cv.survey[[1]])
 )
 
 ###############################################################################
@@ -164,7 +165,7 @@ unlink("FIMS_input_data.csv")
 
 usethis::use_data(data_mile1, overwrite = TRUE)
 rm(
-  check_ASSAMC,
+  check_ASSAMC, cv_2_sd,
   age_data, landings_data, index_data, weightatage_data,
   timingfishery, weightsfishery,
   data_mile1, returnedom,