5_SummarizeResults_SpecialQueries.Rmd

---
title: "Impact Indicators: Special Queries"
author: "Cameryn Brock"
date: "4/21/2022"
output: html_document
editor_options: 
  chunk_output_type: console
---

```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = TRUE)
```

```{r}
library(tidyverse)

year <- "2022"

# read in data for 'other indicators' for all sites and overlaps
sites <- read_csv(paste0(
  "results/FY", str_sub(year, start = 3), "_ImpactIndicators_Other_Sites.csv"))

overlaps <- readRDS(
  paste0("results/FY", str_sub(year, start = 3), "_ImpactIndicators_Other_Overlaps.rds"))

# Define summarizing function
other_summarize <- function(df){
  df %>% 
    summarize(across(
      .cols = c(area_ha, rest_area, population, 
                tstor_woody, tstor_soil, tstor_total, 
                carbon_seq_potl),
      sum, na.rm = TRUE),
      .groups = "keep") 
}

vars <- c(
  "country",
  "ci_divisio",
  "ci_sls_2")
```

## 1. Southern Cross Area Under Restoration

* Reporting levels = country and scape; please also include totals by restoration type
* Restoration Type column must be selected (remove blanks and “not applicable”) 
* Use the CI start date column - should not include anything before 2017 (prior to SC)

```{r, Southern Cross Area Under Restoration}

for (v in seq_along(vars)){
  
  user_group_1 <- vars[v]
  user_group_2 <- "restoratio"
  
  user_groups <- c(user_group_1, user_group_2)
  
  sites_summary <- sites %>% 
    # remove sites before 2017
    filter(ci_start_d > as.Date('2017-01-01'),
           restoratio != "Not Applicable") %>% 
    group_by_at(user_groups) %>% 
    other_summarize() %>% 
    ungroup() 
  
  remove_ids <- sites %>% 
    filter(ci_start_d < as.Date('2017-01-01') | restoratio == "Not Applicable") %>% 
    pluck("ci_id") %>% 
    unique()
  
  # summarize overlaps
  overlaps_summary <- overlaps %>% 
    rowwise() %>% 
    mutate(remove = any(unlist(ci_id) %in% remove_ids)) %>%
    ungroup() %>% 
    filter(!remove == TRUE) %>% 
    dplyr::select((!!as.symbol(user_group_1)), (!!as.symbol(user_group_2)), 
                  area_ha, rest_area, population, tstor_woody, 
                  tstor_soil, tstor_total, carbon_seq_potl) %>% 
    rowwise() %>% 
    # get a list of duplicated values
    mutate(duplicates = list((!!as.symbol(user_group_2))[duplicated((!!as.symbol(user_group_2)))])) %>% 
    ungroup() %>% 
    # remove those with no duplicates
    filter(!duplicates == "character(0)") %>% 
    # multiply values of interest by the number of times its duplicated & make negative 
    rowwise() %>%
    mutate({{ user_group_2 }} := list(unique(duplicates))) %>% 
    unnest({{ user_group_2 }}) %>% 
    ungroup() %>% 
    rowwise() %>% 
    mutate(duplicate_n = sum((!!as.symbol(user_group_2)) == duplicates, na.rm = TRUE)) %>% 
    mutate(across(
      .cols = c(area_ha, rest_area, population, tstor_woody, 
                tstor_soil, tstor_total, carbon_seq_potl),
      ~ .x * -duplicate_n))  %>% 
    
    # do again for second grouping if needed
    mutate(duplicates = list((!!as.symbol(user_group_1))[duplicated((!!as.symbol(user_group_1)))])) %>% 
    ungroup() %>% 
    filter(!duplicates == "character(0)") %>% 
    rowwise() %>%
    mutate({{ user_group_1 }} := unique(duplicates)) %>% 
    #      unnest({{ user_group_1 }}) %>% 
    ungroup() %>% 
    # summarize as you did with sites
    group_by_at(user_groups) %>% 
    other_summarize() %>% 
    ungroup() 
  
  # combine the two data frames and summarize to subtract the overlaps
  corrected_summary <- sites_summary %>% 
    bind_rows(overlaps_summary) %>% 
    group_by_at(user_groups) %>% 
    other_summarize() %>% 
    ungroup() %>% 
    dplyr::select(c(1,2,4))
  
  write_csv(
    corrected_summary,
    file = paste0("results/summaries_", year, "/special_queries/ImpactIndicators", "_", str_to_title(user_group_1),
                  "_", str_to_title(user_group_2), "_restAreaPost2017.csv")
  )
  
}
```

## 2.  Southern Cross Potential Sequestration

* Reporting levels = country and scape; please also include totals by restoration type
* Restoration Type column must be selected (remove blanks and “not applicable”) 
* If restoration type = Agroforestry, Enrichment Planting/Assisted Natural Regeneration, Mangrove Shrub Restoration, Mangrove Tree Restoration, Plantations & Woodlots – Mixed 50/50, Seed Dispersal, Wetland Restoration - Use the tree planting date – should not include anything prior to 2018 or after June 30, 2021 
* If restoration type = Rangeland Restoration, natural regeneration - Use the CI start date - should not include anything prior to 2018 or after June 30, 2021 


```{r, Southern Cross Potential Sequestration}

for (v in seq_along(vars)){
  
  user_group_1 <- vars[v]
  user_group_2 <- "restoratio"
  
  user_groups <- c(user_group_1, user_group_2)
  
  sites_summary <- sites %>% 
    # filter(restoratio != "Not Applicable") %>% 
    # # remove sites as requested
    # mutate(remove = case_when(
    #   tree_plant > as.Date('2018-01-01') & 
    #     tree_plant < as.Date('2021-06-30') ~ "keep",
    #   restoratio %in% c("Rangeland Restoration - Planned Grazing", "Natural Regeneration") &
    #     ci_start_d > as.Date('2018-01-01') & 
    #     ci_start_d < as.Date('2021-06-30') ~ "keep",
    #   T ~ "remove"
    # )) %>% 
    # filter(remove == "keep") %>% 
    # dplyr::select(!remove) %>% 
    group_by_at(user_groups) %>% 
    other_summarize() %>% 
    ungroup() 
  
  remove_ids <- sites %>% 
    # remove sites as requested
    mutate(remove = case_when(
      tree_plant > as.Date('2018-01-01') & 
        tree_plant < as.Date('2021-06-30') ~ "keep",
      restoratio %in% c("Rangeland Restoration - Planned Grazing", "Natural Regeneration") &
        ci_start_d > as.Date('2018-01-01') & 
        ci_start_d < as.Date('2021-06-30') ~ "keep",
      T ~ "remove"
    )) %>% 
    filter(remove == "remove") %>% 
    pluck("ci_id") %>% 
    unique()
  
  # summarize overlaps
  overlaps_summary <- overlaps %>% 
    rowwise() %>% 
    mutate(remove = any(unlist(ci_id) %in% remove_ids)) %>%
    ungroup() %>% 
    filter(!remove == TRUE) %>% 
    dplyr::select((!!as.symbol(user_group_1)), (!!as.symbol(user_group_2)), 
                  area_ha, rest_area, population, tstor_woody, 
                  tstor_soil, tstor_total, carbon_seq_potl) %>% 
    rowwise() %>% 
    # get a list of duplicated values
    mutate(duplicates = list((!!as.symbol(user_group_2))[duplicated((!!as.symbol(user_group_2)))])) %>% 
    ungroup() %>% 
    # remove those with no duplicates
    filter(!duplicates == "character(0)") %>% 
    # multiply values of interest by the number of times its duplicated & make negative 
    rowwise() %>%
    mutate({{ user_group_2 }} := list(unique(duplicates))) %>% 
    unnest({{ user_group_2 }}) %>% 
    ungroup() %>% 
    rowwise() %>% 
    mutate(duplicate_n = sum((!!as.symbol(user_group_2)) == duplicates, na.rm = TRUE)) %>% 
    mutate(across(
      .cols = c(area_ha, rest_area, population, tstor_woody, 
                tstor_soil, tstor_total, carbon_seq_potl),
      ~ .x * -duplicate_n))  %>% 
    
    # do again for second grouping if needed
    mutate(duplicates = list((!!as.symbol(user_group_1))[duplicated((!!as.symbol(user_group_1)))])) %>% 
    ungroup() %>% 
    filter(!duplicates == "character(0)") %>% 
    rowwise() %>%
    mutate({{ user_group_1 }} := unique(duplicates)) %>% 
    #      unnest({{ user_group_1 }}) %>% 
    ungroup() %>% 
    # summarize as you did with sites
    group_by_at(user_groups) %>% 
    other_summarize() %>% 
    ungroup() %>% 
    filter(!restoratio == "Not Applicable")
  
  # combine the two data frames and summarize to subtract the overlaps
  corrected_summary <- sites_summary %>% 
    bind_rows(overlaps_summary) %>% 
    group_by_at(user_groups) %>% 
    other_summarize() %>% 
    ungroup() %>% 
    dplyr::select(c(1,2,9))
  
  write_csv(
    corrected_summary,
    file = paste0("results/summaries_", year, "/special_queries/ImpactIndicators", "_", str_to_title(user_group_1),
                  "_", str_to_title(user_group_2), "_carbonSeqPost2018.csv")
  )
  
}
```

## 3. Southern Cross Marine Protected Area 

* Reporting levels = by country and scape (CI_SLS_2 column)
* Query "Total Marine Protected Area" (square kilometer) 
    + Biome Column = Marine; Terrestrial + Marine 
    + Intervention Type= Protected Area (Note: please do not include Protected Area – Proposed) 
    + Gazettement Date= Should fall within FY21 (July 1 2020 – June 30 2021) OR 
    + OR Improved Management Column = Include all options with “yes” in the selection (ex. “Yes, staff increased”); Remove - Blanks and Not Applicable 
    + CI Division - Do not include anything tagged as “C4O - Blue Nature Alliance” 

* New Marine Protected Area 
    + Biome Column = Marine; Terrestrial + Marine 
    + Intervention Type= Protected Area (Note: please do not include Protected Area – Proposed) 
    + Gazettement Date= Should fall within FY21 (July 1 2020 – June 30 2021) 
    + CI Division - Do not include anything tagged as “C4O - Blue Nature Alliance” 

* Marine Protected Area with Improved Management Activities 
    + Biome Column = Marine; Terrestrial + Marine 
    + Intervention Type = Protected Area (Note: please do not include Protected Area – Proposed) 
    + Improved Management Column = Include all options with “yes” in the selection (ex. “Yes, staff increased”); Remove - Blanks and Not Applicable 
    + CI Division - Do not include anything tagged as “C4O - Blue Nature Alliance” 



```{r, Southern Cross Marine Protected Area}

for (v in seq_along(vars)){
  
  user_group <- vars[v]
  
  sites_summary <- sites %>% 
    filter(
      biome %in% c("Marine", "Terrestrial + Marine"),
      interventi == "Protected Area (National or Regional)",
      gazettemen > as.Date("2020-07-01") &
        gazettemen < as.Date("2021-06-30") | 
        str_detect(improved_m, "Yes")) %>% 
    filter(is.na(ci_divis_1) | ci_divis_1 != "C4O - Blue Nature Alliance") %>% 
    group_by_at(user_group) %>% 
    summarize(
      area_ha = sum(area_ha)) %>% 
    ungroup() 
  
  remove_ids <- sites %>% 
    filter(
      !biome %in% c("Marine", "Terrestrial + Marine") |
        interventi == "Protected Area (National or Regional)" | 
        (is.na(ci_divis_1) | ci_divis_1 != "C4O - Blue Nature Alliance") | 
        (gazettemen > as.Date("2020-07-01") &
           gazettemen < as.Date("2021-06-30") | 
           str_detect(improved_m, "Yes"))) %>% 
    pluck("ci_id") %>% 
    unique()
  
  # summarize overlaps
  overlaps_summary <- overlaps %>% 
    rowwise() %>% 
    mutate(remove = any(unlist(ci_id) %in% remove_ids)) %>%
    ungroup() %>% 
    filter(!remove == TRUE) 
  # 0 overlaps - don't need to correct anything from sites_summary
  # may have to adjust in future years if there are overlaps
  
  corrected_summary <- sites_summary
  
  write_csv(
    corrected_summary,
    file = paste0("results/summaries_", year, "/special_queries/ImpactIndicators", "_", str_to_title(user_group),
                  "_marineProtectedAreas.csv")
  )
  
}
```