### Information
### This script is used to convert the original guilds data to a userfriendly file.
### The input is the Gilden_Info (converted form Excel to a CSV)
### The output is the Gilden_Data, where every species is listed only once per guild.
### The output must be read in to Excel manually to correctly display ä,ö or ü. 

### This script accompanies the 2026 report and dataset produced by the Translational Centre
    # biodiversity conservation, German title: "Ein Datensatz und Toolkit zur Verbesserung der Lebensraumvernetzung"
### Script developed by Johanna Gnägi and Sarah Richman.


# Load packages
library(dplyr)
library(stringr)
library(readr)

# Help function for summarising text elements
clean_collapse <- function(x) {
  x %>%
    na.omit() %>%
    .[. != ""] %>%
    unique() %>%
    str_squish() %>%
    paste(collapse = " ; ")}


# Load data. Add file path and file name between quotation marks 
gilden_orig <- read_csv(
  "",
  col_types = cols(
    seed_mass = col_double()))


# Process Data
gilden_data <- gilden_orig %>%
  
  # source in one field
  mutate(source = str_c(data_year, data_country, data_source, sep = " | ")) %>%
  
  # group and summarize
  group_by(main_guilde, species, tax_id, tax_group) %>%
  summarise(
    species_syn = first(species_syn),
    species_de = first(species_de),
    add_guilds = first(add_guilds),
    
    # mean of numeric variables
    across(
      c(dd_low, hr_low, height, seed_mass, seed_size),
      ~ if (all(is.na(.))) NA else mean(., na.rm = TRUE),
      .names = "{.col}" ),
    
    # summarize text with help function
    dd_info = clean_collapse(dd_info),
    hr_info = clean_collapse(hr_info),
    disp_type = clean_collapse(disp_type),
    density = clean_collapse(density),
    barriers = clean_collapse(barriers),
    source = clean_collapse(source),
    
    .groups = "drop") %>%
  
  # Keep only species with some information
  filter(
    if_any(
      .cols = c(dd_low,hr_low,),
      ~ !is.na(.) & . != ""))


# Export. Fill parenthesis with desired file path and file name 
write_csv()