total_catch by

R/get_total_catch.R

@@ -5,9 +5,9 @@
 #' @param crab_data NULL. Observer measure pot data in the same format as the output of `load_crab_dump()`
 #' @param ft_data NULL. Fish ticket data with fields 'crab_year', 'fishery', 'effort' and whatever variables are specified in by.
 #' @param stock NULL. Character string stock abbreviation: BSSC, WBT, EBT, BBRKC, EAG, WAG, PIGKC, SMBKC, PIBKC, PIRKC, WAIRKC.
-#' @param by NULL. Vector of column names other than crab year, fishery, and group to use as a grouping variable. Must by present in count pot data.
+#' @param by NULL. Vector of column names other than crab year, fishery, and group to use as a grouping variable.
 #' @param units NULL. Unit of measurement, "t" or "lb". Default = "t".
-#' @return Data frame including crab year, fishery, grouping variables, and average weight.
+#' @return Data frame including crab year, fishery, grouping variables, and average weight. Average weights used for expansion to weight are only based on crab year, fishery, and sex. No other 'by' variables are included here, so that the sum of total catch by variables (i.e., sex and shell condition) will equal the total catch by sex.
 #' @examples get_avg_wt(pot_data, crab_data, ft_data, stock = "BBRKC")
 #'
 #' @export
@@ -16,63 +16,139 @@ get_total_catch <- function(pot_data, crab_data, ft_data, stock, by = NULL, unit
 
   if(!(units %in% c("t", "lb"))){stop("Do not recognize the units, use t or lb")}
 
+  by_in <- by
   # make sure fishery and crab_year are not in by
-  by = by[!(by %in% c("fishery", "crab_year", "group"))]
+  by <- by[!(by %in% c("fishery", "crab_year", "group", "sex"))]
+  # 'by' for pot data can only include columns %in% count data
+  by_cp <- by[by %in% names(pot_data)]
+
   if(length(by) == 0){by = NULL}
+  if(length(by_cp) == 0){by_cp = NULL}
+  if(!is.null(by_cp)){
+    if(!(by_cp %in% names(ft_data))){stop(paste0("One or more of ", by_cp, " not in fish ticket data !!"))}
+  }
+  # clean up by data in crab data and save it as a separate object
+  crab_data_tc <- crab_data
+  if("shell" %in% by & lump == T){
+    crab_data_tc %>%
+      filter(!is.na(shell),
+             shell != -9) %>%
+      mutate(shell = case_when(shell %in% c(0:2, 9) ~ "new",
+                               shell %in% c(3:5) ~ "old")) -> crab_data_tc
+  }
+  if("shell" %in% by & lump == F){
+    crab_data_tc %>%
+      filter(!is.na(shell),
+             shell != -9)  -> crab_data_tc
+  }
+  if("legal" %in% by){
+    crab_data_tc %>%
+      filter(!is.na(legal),
+             legal != -9)  -> crab_data_tc
+  }
+  if("sex" %in% by_in){
+    crab_data_tc %>%
+      filter(sex %in% 1:2)  -> crab_data_tc
+  }
 
   # get average wt
   crab_data %>%
     # remove known bad biotwine records
     filter((biotwine_ok %in% c("Y", "-") | is.na(biotwine_ok))) %>%
     # add average weight
     add_calc_wt(., stock = stock, units = units) %>%
-    get_avg_wt(., stock = stock, by = c("group", by), units = units) -> avg_wt
+    get_avg_wt(., stock = stock, by = c("group"), units = units) %>%
+    mutate(sex = case_when(group == "female" ~ 2,
+                           group == "sublegal_male"~ 1,
+                           group == "legal_male"~ 1)) %>%
+    dplyr::select(crab_year, fishery, group, sex, avg_wt) -> avg_wt
 
   # get average weight of all years by fishery
   avg_wt %>%
     dplyr::select(-crab_year) %>%
     mutate(fishery = substring(fishery, 1, 2)) %>%
     rename(fish_type = fishery) %>%
     group_by_at(1:(ncol(.) - 1)) %>%
-    summarise(avg_wt_mean = mean(avg_wt)) %>%
-    ungroup -> avg_wt_fish_type
+    summarise(avg_wt_mean = mean(avg_wt), .groups = "drop") -> avg_wt_fish_type
+
+  # get percent of crab measured in by variables
+  crab_data_tc %>%
+    group_by_at(c("crab_year", "fishery", "group", by)) %>%
+    count() %>%
+    group_by_at(c("crab_year", "fishery", "group", by)) %>%
+    group_by(crab_year, fishery, group) %>%
+    mutate(tot = sum(n)) %>% ungroup %>%
+    mutate(prop = n / tot) %>%
+    dplyr::select("crab_year", "fishery", "group", all_of(by), "prop") %>%
+    # expand out to all crab year fishery and group combos
+    right_join(expand_grid(distinct(., crab_year, fishery),
+                          group = c("female", "sublegal_male", "legal_male")),
+               by = join_by(crab_year, fishery, group)) %>%
+    replace_na(list(prop = 0)) %>%
+    # add sex
+    mutate(sex = case_when(group == "female" ~ 2,
+                           group == "sublegal_male"~ 1,
+                           group == "legal_male"~ 1)) %>%
+    dplyr::select(crab_year, fishery, group, sex, all_of(by), prop) -> prop_by
 
   # get observer cpue
   pot_data %>%
     # remove known bad biotwine records
     filter((biotwine_ok %in% c("Y", "-") | is.na(biotwine_ok))) %>%
-    group_by_at(c("crab_year", "fishery", by)) %>%
+    group_by_at(c("crab_year", "fishery", by_cp)) %>%
     summarise(female = sum(female, na.rm = T),
               sublegal_male = sum(sublegal, na.rm = T),
               legal_male = sum(tot_legal, na.rm = T),
-              obs_effort = n()) %>% ungroup%>%
+              obs_effort = n(), .groups = "drop") %>%
     # pivot to long format
     pivot_longer(c(female, sublegal_male, legal_male), names_to = "group", values_to = "count") %>%
-    # cpue
-    mutate(cpue = count / obs_effort) -> obs_cpue
+    replace_na(list(female = 0, sublegal_male = 0, legal_male = 0)) %>%
+    # add sex
+    mutate(sex = case_when(group == "female" ~ 2,
+                           group == "sublegal_male"~ 1,
+                           group == "legal_male"~ 1)) -> obs_count
 
   # summarise fish ticket data
   ft_data %>%
-    group_by_at(c("crab_year", "fishery", by)) %>%
-    summarise(effort = sum(effort, na.rm = T)) %>% ungroup -> ft_data
+    group_by_at(c("crab_year", "fishery", by_cp)) %>%
+    summarise(ft_effort = sum(effort, na.rm = T), .groups = "drop") -> ft_data
 
-  # get total catch
-  if(!is.null(by)) {
-    if(!(by %in% names(ft_data))){stop(paste0(by, " not present in fish ticket data"))}
-  }
   # join to effort and avg wt
-  obs_cpue %>%
-    left_join(ft_data, by = c("crab_year", "fishery", by)) %>%
-    left_join(avg_wt, by = c("crab_year", "fishery", "group", by)) %>%
+  obs_count %>%
+    left_join(ft_data, by = c("crab_year", "fishery", by_cp)) %>%
+    full_join(prop_by, by = c("crab_year", "fishery", "group", "sex", by_cp)) %>%
+    left_join(avg_wt, by = c("crab_year", "fishery", "group", "sex")) %>%
     # fill in average weight when missing if available from another year
     mutate(fish_type = substring(fishery, 1, 2)) %>%
-    left_join(avg_wt_fish_type, by = c("fish_type", "group", by)) %>%
+    left_join(avg_wt_fish_type, by = c("fish_type", "group", "sex")) %>%
     mutate(avg_wt = ifelse(is.na(avg_wt), avg_wt_mean, avg_wt)) %>%
     dplyr::select(-fish_type, -avg_wt_mean) %>%
+    # fill in prop if by is NULL
+    {if(is.null(by)) {replace_na(., list(prop = 1))} else{.}} %>%
+    # scale count
+    # add observer cpue
+    mutate(count = count * prop,
+           cpue = count / obs_effort) %>%
     # scale to total catch
-    mutate(total_catch_n = cpue * effort,
-           total_catch_wt = cpue * effort * avg_wt)  %>%
-    dplyr::select(crab_year, fishery, by, group, count, obs_effort, effort, cpue, avg_wt, total_catch_n, total_catch_wt) -> out
+    mutate(total_catch_n = ifelse(count != 0, cpue * ft_effort, 0),
+           total_catch_wt = ifelse(count != 0, cpue * ft_effort * avg_wt, 0))  %>%
+
+    dplyr::select(crab_year, fishery, group, all_of(by_in), count, obs_effort, ft_effort, cpue, avg_wt, total_catch_n, total_catch_wt) -> out
+
+  if(("sex" %in% by_in) & !("group" %in% by_in)) {
+
+    out %>%
+      group_by_at(c("crab_year", "fishery", "sex", by)) %>%
+      summarise(count_s = sum(count),
+                obs_effort = mean(obs_effort),
+                ft_effort = mean(ft_effort),
+                cpue = count_s / obs_effort,
+                avg_wt = weighted.mean(avg_wt, ifelse(count == 0, 1, count)),
+                total_catch_n = cpue * ft_effort,
+                total_catch_wt = cpue * ft_effort * avg_wt, .groups = "drop") %>%
+      rename(count = count_s) -> out
+
+  }
 
   return(out)