Gaitlyn malone noaa patch 1

ODM-MH-Analysis_ready/Closures/Closure_Reopening.R → ODM-MH-Analysis_ready/Closures/Closure_tests/Closure_Reopening.R

@@ -4,25 +4,26 @@
 librarian::shelf(here, tidyverse, lubridate, dplyr, tidyr, neatRanges, splitstackshape)
 
 # Read in MH Data log ####
-# mh_data_log <- readRDS(here("ODM-MH-Data_log", "data", "results", "MH_DL_2023Nov13.RDS"))
+ mh_data_log <- readRDS(here("ODM-MH-Data_log", "data", "results", "MH_DL_2024Jul12.RDS"))
 # 
 # # Select species and region ####
-# spp <- 'PORGY, RED'
-# region <- 'SOUTH ATLANTIC'
+ spp <- 'SNAPPER, RED'
+ region <- 'GULF OF MEXICO'
 
 # Seasonal Closures ####
 # Filter dataset to seasonal closure regulations for the specified species and region
-Seasonal_closures <- mh_spp_closure %>%
+Seasonal_closures <- mh_data_log %>%
   filter(MANAGEMENT_TYPE_USE == 'CLOSURE', MANAGEMENT_STATUS_USE == 'SEASONAL', NEVER_IMPLEMENTED == 0, REG_REMOVED == 0)
 
 # Create date_sequence to expand dates between EFFECTIVE_DATE and END_DATE2 
 expand_seasonal <- Seasonal_closures %>%
+  filter(EFFECTIVE_DATE <= END_DATE2) %>%
   mutate(date_sequence = map2(EFFECTIVE_DATE, END_DATE2, seq, by = "days")) %>%
   unnest(date_sequence)
 
 # Add START_YEAR_expand and END_YEAR_expand to extract the year from the date_sequence field
 expand_seasonal_year <- expand_seasonal %>%
-  #rowwise() %>%
+  rowwise() %>%
   mutate(START_YEAR_expand = year(date_sequence),
          END_YEAR_expand = year(date_sequence))
 
@@ -227,7 +228,7 @@ chk <- closure_all %>%
 # Join in Long Form
 # Join Closure Types Together ####
 # Combine closure data frames
-Combined_closures <- bind_rows(add_closure_value_seasonal, add_closure_value_weekly, add_closure_value_one)
+Combined_closures <- bind_rows(add_closure_value_seasonal, add_closure_value_monthly, add_closure_value_one)
 
 # Sort the data frame by date_sequence
 sort_closures <- Combined_closures %>%
@@ -252,7 +253,7 @@ closures_remove_fields <- closures_remove_multi %>%
          -START_DAY_OF_WEEK_EXPAND, -END_DAY_OF_WEEK_EXPAND, -Multi_expand)
 
 # Create EFFECTIVE_YEAR field to indicate the effective year of the closure
-closures_effective_year <- closures_remove_fields %>%
+closures_effective_year <- closures_remove_multi %>%
   mutate(EFFECTIVE_YEAR = year(date_sequence))
 
 # Create EFFECTIVE_MONTH field to indicate the effective month of the closure

ODM-MH-Analysis_ready/Closures/Main_closure_prep.R

@@ -1,20 +1,34 @@
-# Process all regulations related to closures
+#### Closures ####
+# Code creates a comprehensive, expanded time series of fishery status (OPEN/CLOSE)
+# by incorporating all closure-related regulations.
+# The first output from this code indicates the days open per year by fishery (commercial/recreational)
+# The second output from this code displays the continuous OPEN/CLOSE periods with start/end dates
 
 # Load packages ####
 librarian::shelf(here, tidyverse, lubridate, dplyr, tidyr, neatRanges, splitstackshape)
 
-# Read in MH Data Log
-mh_data_log <- readRDS(here("ODM-MH-Data_log", "data", "results", "MH_DL_2025Mar28.RDS"))
+# Read in MH Data Log ####
+mh_data_log <- readRDS(here("ODM-MH-Data_log", "data", "results", "MH_DL_2025Sep10.RDS"))
 
-# Select species and region
+# Select species and region of interest for the closure time series ####
 spp <- 'SNAPPER, RED'
-region <- 'GULF OF MEXICO'
+region <- 'SOUTH ATLANTIC'
 
-# Function to expand dates based on management status
+# Function to expand dates based on management status ####
+# Function expands records for a given MANAGEMENT_TYPE_USE into a daily date sequence with
+# the status of the fishery (OPEN/CLOSED) indicated for each day based on the MANAGEMENT_STATUS_USE 
+# (ONCE/SEASONAL/RECURRING) of the applicable regulations
 source(here("ODM-MH-Analysis_ready", "Closures", "func_expand_status.R"))
 
-# filter for species and region (all closure related records)
-# Recode management types to have a value
+# Filter and reclassify closure-related records ####
+# Retain only species and region of interest and management types that impact the status of the fishery
+# Provide management types that impact fishery status with a VALUE of OPEN or CLOSE
+# FISHING SEASON/FISHING YEAR implies that the fishery is OPEN during the defined window
+# PROHIBITED SALE AND PURCHASE 
+#    - If FLAG = YES -> OPEN (sale is allowed)
+#    - If FLAG = NO -> CLOSE (sale is not allowed)
+# PROHIBITED SPECIES implies the fishery is closed as species cannot be retained
+# CLOSURE VALUE already indicates fishery status
 mh_spp_closure <- mh_data_log %>%
   filter(COMMON_NAME_USE == spp,
          REGION == region,
@@ -28,39 +42,46 @@ mh_spp_closure <- mh_data_log %>%
                            TRUE ~ VALUE)) %>%
   arrange(SECTOR_USE, START_DATE2)
 
+# Check to see which MANAGEMENT_TYPE_USE are present after filtering
 unique(mh_spp_closure$MANAGEMENT_TYPE_USE)
 
-# Fishing Year
-chk <- filter(mh_spp_closure, MANAGEMENT_TYPE_USE == "FISHING YEAR")
+# Date expansion by MANAGEMENT_TYPE_USE ####
+# Each section expands the raw rules for the applicable MANAGEMENT_TYPE_USE into a 
+# series by date with the associated status (OPEN/CLOSE) and START/END TIME
+
+# Fishing Year expansion
+unique(select(filter(mh_spp_closure, MANAGEMENT_TYPE_USE == "FISHING YEAR"), MANAGEMENT_STATUS_USE))
 spp_year <- expand_status(mh_spp_closure, "FISHING YEAR")
 
-# Fishing Season - still needs some work if not entered consistently
-chk <- filter(mh_spp_closure, MANAGEMENT_TYPE_USE == "FISHING SEASON")
+# Fishing Season expansion 
+# Portion may require some adjustment if fishing year has not been documented consistently in the raw data
+unique(select(filter(mh_spp_closure, MANAGEMENT_TYPE_USE == "FISHING SEASON"), MANAGEMENT_STATUS_USE))
 spp_season <- expand_status(mh_spp_closure, "FISHING SEASON")
 
-# Closure
+# Closure expansion
 unique(select(filter(mh_spp_closure, MANAGEMENT_TYPE_USE == "CLOSURE"), MANAGEMENT_STATUS_USE))
 spp_closures <- expand_status(mh_spp_closure, "CLOSURE")
 
-# Prohibited sale and purchase
+# Prohibited Sale and Purchase expansion
 unique(select(filter(mh_spp_closure, MANAGEMENT_TYPE_USE == "PROHIBITED SALE AND PURCHASE"), MANAGEMENT_STATUS_USE))
 spp_prohibited_sale <- expand_status(mh_spp_closure, "PROHIBITED SALE AND PURCHASE")
 
-# Prohibited species
+# Prohibited Species expansion
 unique(select(filter(mh_spp_closure, MANAGEMENT_TYPE_USE == "PROHIBITED SPECIES"), MANAGEMENT_STATUS_USE))
 spp_prohibited_spp <- expand_status(mh_spp_closure, "PROHIBITED SPECIES")
 
-# Combine all management types that refer to closures
-# For now just comment out mtypes that do not apply
+# Join all expanded series into one time series ####
+# For days with multiple statuses based on multiple regulations, select the most recent FR Citation
 spp_closure_story <- spp_year %>%
   rename(FR_CITATION_year = "FR_CITATION",
          TIME_year = "TIME",
          VALUE_year = "VALUE") %>%
   select(FMP, COMMON_NAME_USE, REGION, ZONE_USE, SECTOR_USE, SUBSECTOR_USE, date_sequence, FR_CITATION_year, TIME_year, VALUE_year) %>%
+  # Comment out MANAGEMENT_TYPE_USE that do not apply to your filter criteria
   full_join(spp_season %>%
-              rename(FR_CITATION_season = "FR_CITATION",
+            rename(FR_CITATION_season = "FR_CITATION",
                      TIME_season = "TIME",
-                     VALUE_season = "VALUE") %>%
+                    VALUE_season = "VALUE") %>%
               select(FMP, COMMON_NAME_USE, REGION, ZONE_USE, SECTOR_USE, SUBSECTOR_USE, date_sequence, FR_CITATION_season, TIME_season, VALUE_season),
             by = join_by(FMP, COMMON_NAME_USE, REGION, ZONE_USE, SECTOR_USE, SUBSECTOR_USE,  date_sequence)) %>%
   full_join(spp_closures %>%
@@ -69,57 +90,67 @@ spp_closure_story <- spp_year %>%
                      VALUE_close = "VALUE") %>%
               select(FMP, COMMON_NAME_USE, REGION, ZONE_USE, SECTOR_USE, SUBSECTOR_USE, date_sequence, FR_CITATION_close, TIME_close, VALUE_close),
             by = join_by(FMP, COMMON_NAME_USE, REGION, ZONE_USE, SECTOR_USE, SUBSECTOR_USE,  date_sequence)) %>%
- # full_join(spp_prohibited_sale %>%
-   #          rename(FR_CITATION_sale = "FR_CITATION",
-    #                TIME_sale = "TIME", 
-    #                 VALUE_sale = "VALUE") %>%
-     #         select(FMP, COMMON_NAME_USE, REGION, ZONE_USE, SECTOR_USE, SUBSECTOR_USE, date_sequence, FR_CITATION_sale, VALUE_sale),
-      #      by = join_by(FMP, COMMON_NAME_USE, REGION, ZONE_USE, SECTOR_USE, SUBSECTOR_USE,  date_sequence)) %>%
- # full_join(spp_prohibited_spp %>%
-  #            rename(FR_CITATION_spp = "FR_CITATION",
-  #                   TIME_spp = "TIME",
-   #                  VALUE_spp = "VALUE") %>%
-    #          select(FMP, COMMON_NAME_USE, REGION, ZONE_USE, SECTOR_USE, SUBSECTOR_USE, date_sequence, FR_CITATION_spp, VALUE_spp),
-     #         by = join_by(FMP, COMMON_NAME_USE, REGION, ZONE_USE, SECTOR_USE, SUBSECTOR_USE, date_sequence)) %>%
+  full_join(spp_prohibited_sale %>%
+             rename(FR_CITATION_sale = "FR_CITATION",
+                    TIME_sale = "TIME", 
+                     VALUE_sale = "VALUE") %>%
+              select(FMP, COMMON_NAME_USE, REGION, ZONE_USE, SECTOR_USE, SUBSECTOR_USE, date_sequence, FR_CITATION_sale, VALUE_sale),
+            by = join_by(FMP, COMMON_NAME_USE, REGION, ZONE_USE, SECTOR_USE, SUBSECTOR_USE,  date_sequence)) %>%
+  full_join(spp_prohibited_spp %>%
+              rename(FR_CITATION_spp = "FR_CITATION",
+                     TIME_spp = "TIME",
+                     VALUE_spp = "VALUE") %>%
+              select(FMP, COMMON_NAME_USE, REGION, ZONE_USE, SECTOR_USE, SUBSECTOR_USE, date_sequence, FR_CITATION_spp, VALUE_spp),
+              by = join_by(FMP, COMMON_NAME_USE, REGION, ZONE_USE, SECTOR_USE, SUBSECTOR_USE, date_sequence)) %>%
   arrange(SECTOR_USE, SUBSECTOR_USE, ZONE_USE, date_sequence) %>%
-  # Select the most recent FR CITATION
+  # Select the most recent FR CITATION and retain its TIME/VALUE
   mutate(FR_CITATION = pmax(FR_CITATION_year, 
                             FR_CITATION_season, 
                             FR_CITATION_close, 
-                           # FR_CITATION_sale, 
-                            #FR_CITATION_spp,
+                            FR_CITATION_sale, 
+                            FR_CITATION_spp,
                             na.rm = T),
          TIME = case_when(FR_CITATION == FR_CITATION_close ~ TIME_close,
-                          #FR_CITATION == FR_CITATION_sale ~ TIME_sale,
-                          #FR_CITATION == FR_CITATION_spp ~ TIME_spp,
-                          FR_CITATION == FR_CITATION_season ~ TIME_season,
+                          FR_CITATION == FR_CITATION_sale ~ TIME_sale,
+                          FR_CITATION == FR_CITATION_spp ~ TIME_spp,
+                         FR_CITATION == FR_CITATION_season ~ TIME_season,
                           FR_CITATION == FR_CITATION_year ~ TIME_year)) %>%
-  # Select the fishery status (open/closed) that applies to the most recent FR
   mutate(VALUE = case_when(FR_CITATION == FR_CITATION_close ~ VALUE_close,
-                           #FR_CITATION == FR_CITATION_sale ~ VALUE_sale,
-                           #FR_CITATION == FR_CITATION_spp ~ VALUE_spp,
+                           FR_CITATION == FR_CITATION_sale ~ VALUE_sale,
+                           FR_CITATION == FR_CITATION_spp ~ VALUE_spp,
                            FR_CITATION == FR_CITATION_season ~ VALUE_season,
                            FR_CITATION == FR_CITATION_year ~ VALUE_year)) %>%
-  select(-FR_CITATION_year, -TIME_year, -VALUE_year, -FR_CITATION_close, -TIME_close, -VALUE_close, -FR_CITATION_season, -TIME_season, -VALUE_season) %>%
-  # Create day_calc to indicate cases where the start/end time is midday 
-  # midday start/end times should be counted toward both open and closed periods
+  # Remove management type-specific columns now that a single FR/TIME/VALUE has been selected for each day
+  select(-FR_CITATION_year, -TIME_year, -VALUE_year, -FR_CITATION_close, -TIME_close, -VALUE_close, -FR_CITATION_sale, -VALUE_sale, -TIME_sale, -FR_CITATION_season, -VALUE_season, -TIME_season) %>%
+  # How to handle half days
+  # Some FR notices take effect midday or later. To correctly count days for annual tallies, treat noon/6pm
+  # TIMEs as half-days) (0.5)
+  # This ensures that a midday start/end time contributes to both OPEN and CLOSED period calculations
   mutate(day_calc = case_when(TIME == "06:00:00 PM" ~ 0.5,
                               TIME == "12:00:00 PM" ~ 0.5,
                               TRUE ~ 1))
 
-# Summarize the open and closed periods for each year
+# Summarize the open and closed periods for each year ####
+# Fill in any mission VALUE as OPEN (assumes that any window without an applicable rule is open)
 summ_spp_closures <- spp_closure_story %>%
   mutate(YEAR = year(date_sequence),
          VALUE = as.character(VALUE),
          VALUE = if_else(is.na(VALUE), "OPEN", VALUE)) %>%
   filter(VALUE %in% c("OPEN", "CLOSE"))
 
+# Create half-day complements ####
+# When a day is counted as 0.5 under one status (OPEN/CLOSED), add a complementary 0.5 day to 
+# the opposite status so that the calendar day sums to 1 acorss OPEN/CLOSE
 summ_spp_closures2 <- summ_spp_closures %>%
   filter(day_calc == 0.5) %>%
   mutate(VALUE = if_else(VALUE == "OPEN", "CLOSE", "OPEN"))
 
+# Combine whole-day and half-day rows for final daily time series
 spp_closure_combined <- bind_rows(summ_spp_closures, summ_spp_closures2)
 
+# Collapse individual days into continuous periods within each year ####
+# Group consecutive days with the same VALUE into periods and compute total days within the period (ndays)
+# and list the start/end as the first/last days within the period
 spp_close_sum <- spp_closure_combined %>%
   group_by(FMP, COMMON_NAME_USE, REGION, ZONE_USE, SECTOR_USE, SUBSECTOR_USE, YEAR) %>% 
   arrange(FMP, COMMON_NAME_USE, REGION, ZONE_USE, SECTOR_USE, SUBSECTOR_USE, date_sequence, TIME) %>%
@@ -134,25 +165,10 @@ spp_close_sum <- spp_closure_combined %>%
   arrange(FMP, COMMON_NAME_USE, REGION, ZONE_USE, SECTOR_USE, SUBSECTOR_USE, YEAR, start)
 
 
-# Summarize the open and closed periods for each year
-#summ_spp_closures <- spp_closure_story %>%
- # mutate(YEAR = year(date_sequence)) %>%
-  #group_by(FMP, COMMON_NAME_USE, REGION, ZONE_USE, SECTOR_USE, SUBSECTOR_USE, YEAR) %>% 
-  #arrange(FMP, COMMON_NAME_USE, REGION, ZONE_USE, SECTOR_USE, SUBSECTOR_USE, date_sequence, TIME) %>%
-  #mutate(change = VALUE != lag(VALUE, default = first(VALUE)) |
-   #        YEAR != lag(YEAR, default = first(YEAR))) %>%
-  #mutate(group = cumsum(change)) %>%
-  #group_by(FMP, COMMON_NAME_USE, REGION, ZONE_USE, SECTOR_USE, SUBSECTOR_USE, YEAR, VALUE, group) %>%
-  #summarize(ndays = sum(day_calc, na.rm = TRUE),
-   #         start = min(date_sequence),
-    #        end = max(date_sequence),
-     #       .groups = 'drop') %>%
-  #mutate(VALUE = case_when(is.na(VALUE) ~ "OPEN",
-   #                        TRUE ~ VALUE)) %>%
-  #arrange(FMP, COMMON_NAME_USE, REGION, ZONE_USE, SECTOR_USE, SUBSECTOR_USE, YEAR)
-
-# Identify full range of years per group
-all_years_by_group <- spp_close %>%
+# Identify full range of years per group ####
+# Determine the min/max year of regulations documented and expand to all years between so that even years with zero regulations 
+# are accounted for in the expansion
+all_years_by_group <- spp_close_sum %>%
   distinct(FMP, COMMON_NAME_USE, REGION, ZONE_USE, SECTOR_USE, SUBSECTOR_USE, YEAR, VALUE) %>%
   group_by(FMP, COMMON_NAME_USE, REGION, ZONE_USE, SECTOR_USE, SUBSECTOR_USE) %>%
   summarize(min_year = min(YEAR), max_year = max(YEAR), .groups = "drop") %>%
@@ -161,18 +177,23 @@ all_years_by_group <- spp_close %>%
   unnest(YEAR) %>%
   select(-min_year, -max_year)
 
-# Summarize the number of days open per year
-summ_days_open <- spp_close %>%
+# Summarize the number of days open per year ####
+# Collapse to annual totals of OPEN days per fishery grouping
+summ_days_open <- spp_close_sum %>%
   filter(VALUE == "OPEN") %>%
   group_by(FMP, COMMON_NAME_USE, REGION, ZONE_USE, SECTOR_USE, SUBSECTOR_USE, YEAR) %>%
   summarize(Days_Open = sum(ndays), .groups = "drop") %>%
   arrange(FMP, COMMON_NAME_USE, REGION, ZONE_USE, SECTOR_USE, SUBSECTOR_USE, YEAR)
 
+# Ensure every year in the full span is represented
+# Fill in missing years as zero
 open_by_year <- all_years_by_group %>%
   left_join(summ_days_open, by = c("FMP", "COMMON_NAME_USE", "REGION", "ZONE_USE", "SECTOR_USE", "SUBSECTOR_USE", "YEAR")) %>%
   mutate(Days_Open = replace_na(Days_Open, 0)) %>%
   arrange(FMP, COMMON_NAME_USE, REGION, ZONE_USE, SECTOR_USE, SUBSECTOR_USE, YEAR)
 
-# Save closure output as RDS file
-saveRDS(open_by_year, here("ODM-MH-Analysis_ready", "Closures", "Closure_outputs", paste0('open_by_year', format(Sys.Date(), "%Y%b%d"), '.RDS')))
+# Save annual open days as RDS file
+saveRDS(open_by_year, here("ODM-MH-Analysis_ready", "Closures", "Closure_outputs", paste0('Open_', spp, format(Sys.Date(), "%Y%b%d"), '.RDS')))
 
+# Save closure periods output as RDS file
+saveRDS(spp_close_sum, here("ODM-MH-Analysis_ready", "Closures", "Closure_outputs", paste0('Close_', spp, format(Sys.Date(), "%Y%b%d"), '.RDS')))