Fetch a series as it appeared on a given vintage date

Returns the values that were available in FRED on date, before any subsequent revisions. This is the standard real-time data access pattern: set realtime_start = realtime_end = date. Useful for backtesting forecasting models against the data that was actually available at the time, not the revised series we see today.

Usage

fred_as_of(
  series_id,
  date,
  from = NULL,
  to = NULL,
  units = "lin",
  frequency = NULL,
  aggregation = "avg",
  cache = TRUE
)

Arguments

series_id: Character. One or more FRED series IDs.
date: Character or Date. The vintage date ("YYYY-MM-DD").
from, to: Optional observation date range.
units: Character. Raw FRED units code. Default "lin".
frequency, aggregation: Optional frequency aggregation arguments (see fred_series()).
cache: Logical. Cache results locally. Default TRUE.

Value

A fred_tbl with columns date, series_id, value, realtime_start, realtime_end.

Details

Underneath, this calls the series/observations endpoint with the realtime parameters set. Results are cached separately from the default (latest-vintage) cache, so calling fred_series("GDP") and fred_as_of("GDP", "2020-01-15") keep distinct cache entries.

Examples

# \donttest{
op <- options(fred.cache_dir = tempdir())
# GDP as it looked on 1 March 2020
gdp_2020 <- fred_as_of("GDP", "2020-03-01")
#> Error in fred_get_key(): No FRED API key found.
#> ℹ Set one with `fred_set_key()` or the `FRED_API_KEY` environment variable.
#> ℹ Register for a free key at <https://fredaccount.stlouisfed.org/apikeys>.
options(op)
# }