# FDIC API ----------------------------------------------------------------
# Pull financial information from the FDIC API using package: fdicdata
# Use ?fdicdata or check online pdf documentation for functionality
# Use dataTaxonomy() and search function to explore available information 

# The script relies on the getFinancials() function within fdicdata and
# creates a custom function getFinancialsForAllBanks() to retrieve multiple
# variables for multiple banks across a chosen time frame

# 

# Author: Wade Litt

# Imports -----------------------------------------------------------------
install.packages("fdicdata")
library(fdicdata)
library(dplyr)

# Available Data
taxonomy <- dataTaxonomy("financial")
taxonomy <- dataTaxonomy("institution")

# # Check a single call to ensure the API is responding correctly; example:
# getFinancials(CERT, desired_variables, # records to retrieve, IDRSSD = FALSE, time_range)
getFinancials(3510, c("DRLNLS"), limit=10, FALSE, range = c("2015-01-01","2022-01-01") )


# Bank Selection ----------------------------------------------------------

# Choose population of banks to work with
# Add NAME== or CERT== or IDRSSD== to filter if one specific bank is desired
# Search for specific banks at very bottom of script
df_banks_all <- getInstitutionsAll()
df_banks_selection <- df_banks_all %>% 
  rename(IDRSSD = FED_RSSD) %>% 
  filter(ACTIVE==1,                  # Only active banks 
         ASSET > 1000000000) %>%     # Asset filter (in actual dollars)
  select(NAME, IDRSSD, CERT, ASSET, DATEUPDT, everything()) # reorder columns
bank_ids <- df_banks_selection$CERT  

# create secondary list of banks to join with financial data later
df_banks <- df_banks_selection %>% 
  select(NAME, CITY, STNAME, IDRSSD, CERT) 



# Gather Financial Data ---------------------------------------------------
# Create function to gather financial data from all banks selected above
getFinancialsForAllBanks <- function(bank_ids, metrics, start_date, end_date, limit = 10000) {
  all_data <- data.frame()  # Initialize an empty data frame to store results
  total_banks <- length(bank_ids)
  message(paste("Fetching data... it actually can takes few mins"))
  for (i in seq_along(bank_ids)) {
    bank_id <- bank_ids[i]
    tryCatch({
      # Fetch financial data for the given range, specified below
      temp_data <- getFinancials(bank_id, metrics = metrics, limit = limit, IDRSSD = FALSE, range = c(start_date, end_date))
      # Bind this bank's data to the overall dataframe
      all_data <- rbind(all_data, temp_data)
    }, error = function(e) {
      message(paste("Error fetching data for bank ID", bank_id, ":", e$message))
    })
  }
  
  return(all_data)
}

# !!!!!!! MAIN SELECTION PARAMETERS !!!!!!!!!
# Define metrics and date range
metrics <- c("DRLNLS", "ASSET", "NETINC")  # Defaulted to DRLNLS: Total Loan and Lease Charge-Offs (i.e., loan losses)
start_date <- "2019-01-01"
end_date <- "*"  # Specify an end date or use "*" to run until most recent data
# !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

# Fetch financial data for all banks (using the function created above and parameters
# specified directly above)
df_bank_data <- getFinancialsForAllBanks(bank_ids, metrics, start_date, end_date) %>% 
  left_join(df_banks) %>% 
  select(IDRSSD, CERT, NAME, CITY, STNAME, DATE, everything())






# Extras ------------------------------------------------------------------






# Search for a Specific Bank ----------------------------------------------
searchBankByName <- function(data, search_term) {
  # Convert search term to lowercase to make the search case-insensitive
  search_term <- tolower(search_term)
  
  # Filter data where NAME contains the search_term, case-insensitively
  filtered_data <- data %>%
    filter(grepl(search_term, tolower(NAME)))  # Use grepl for partial matching
  
  return(filtered_data)
}

# Search for banks with last parameter 
df_bank_search <- searchBankByName(df_banks_selection, "Park")