A function to pull multiple data series from FRED

We first want to be able to download time series data from the Frederal Reserve Economic Database (FRED). This can be done using the fredr package that we already included in the above library-statements.

# load fredr library
library(fredr)

In order to be able to download data from FRED, you have to go to https://fred.stlouisfed.org/docs/api/api_key.html. Create an account and request an API key for downloading data. Then insert your API key into the fredr_set_key code line.

The function get_fred_data receives as inputs two arrays that need to have the same length as well as three scalar values. The array series contains the ID of each series that should be downloaded. You can find the ID on the respective series website, in the case of GDP per capita this is this one.. The array names contains the respective names under which the series should be stored in our final data structure. Finally, we need to pass the starting date (start_date), the ending date (end_date), and the frequency (freq) to the function. Note that the starting and end date as well as the frequency need to be identical for all series that should be downloaded.

The function then successively uses the function fredr to download each of the data series and plug them together into one data structure. In addition, the data structure will contain a column with the respective dates to which the data refer.

# In order to be able to download data from FRED, you have to go to
#     https://fred.stlouisfed.org/docs/api/api_key.html
# Create an account and request an API key for downloading data. Then 
# uncomment the statement above and insert your API key.
# All data series available in FRED can be investigated at 
#     https://fred.stlouisfed.org.

#fredr_set_key("type-you-key")

# a function to pull multiple data series from FRED
get_fred_data <- function(series, names, start_date, end_date, freq) {
  
  # extract first data series
  tdat   <- fredr(series_id = series[1],
                  observation_start = as.Date(start_date),
                  observation_end = as.Date(end_date),
                  frequency = freq
  )
  
  # start a new data frame and add first series
  data <- data.frame(date = tdat$date)
  data[names[1]] <- tdat$value
  
  # now iterate over all series and add them
  for(i in 2:length(series)) {
    print(paste("Loading series", series[i]))
    tdat   <- fredr(series_id = series[i],
                    observation_start = as.Date(start_date),
                    observation_end = as.Date(end_date),
                    frequency = freq
    )
    data[names[i]] <- tdat$value
  }
  
  if(freq == "q") {
    data["quarter"] <- c(1:length(unlist(data[names[1]]))) 
  } else if(freq == "a") {
    data["year"] <- c(1:length(unlist(data[names[1]]))) 
  } else {
    data["month"] <- c(1:length(unlist(data[names[1]]))) 
  }
  
  # return the final data frame
  return(data)
}