Skip to contents

Fetch CoinGecko history by numeric ID (incl. partial survivorship-bias correction)

Source: R/cg_recover.R
cg_history_by_id.Rd

Companion to cg_history() that addresses coins by their numeric CoinGecko ID instead of their slug. Useful in two scenarios:

Usage

cg_history_by_id(
  ids = NULL,
  what = c("price", "market_cap", "ohlc"),
  vs_currency = "usd",
  start_date = NULL,
  end_date = NULL,
  coin_list = NULL,
  sleep = 0.6,
  wait = 60,
  max_retries = 3,
  quiet = FALSE,
  finalWait = FALSE,
  date_convention = c("end_of_day", "raw")
)

Arguments

ids

Integer vector of numeric IDs to fetch. Default NULL -> uses cg_list()$id (active universe). To extend coverage to delisted coins, supply the union of historically-observed IDs from your accumulated snapshots.

what

Subset of streams to fetch. Any combination of "price", "market_cap", and "ohlc". Default all three.

vs_currency

Quote currency, default "usd".

start_date, end_date

Client-side date filter applied after fetch. NULL returns full history.

coin_list

Optional cg_list() output used to join slug / name / symbol onto recovered rows for coins still in the active universe. If NULL, calls cg_list() automatically. Set to FALSE to skip the join (rows then have only id).

sleep, wait, max_retries

Rate-limit knobs. Defaults 0.6 / 60 / 3 match cg_history().

quiet

If FALSE, prints a progress bar.

finalWait

Sleep 60 s after the last call (mirrors crypto_history()).

date_convention

Either "end_of_day" (the default) or "raw". See cg_history() for the explanation – this argument applies the same -1 day shift to midnight-UTC ticks so dates align with the CMC / CRSP / Compustat end-of-day convention.

Value

Tibble with one row per (id, date) using crypto2-compatible column names. Columns:

id

CoinGecko numeric id (always populated).

slug, name, symbol

Coin identifiers – NA for ids no longer in the active universe (i.e. delisted on the slug side).

timestamp

POSIXct UTC midnight of the trading day.

ref_cur_id, ref_cur_name

Quote currency.

open, high, low, close, volume, market_cap

Daily values.

Details

  1. Coins that have been delisted and whose slug no longer resolves.

  2. Cronjob-style accumulation: when you persist cg_list() snapshots over time, the union of all numeric IDs ever observed is the survivorship-bias-corrected universe. cg_history_by_id() lets you refetch each historical ID directly without needing its current slug to still resolve.

Important caveats – please read:

  • The numeric-ID space is sparse, not dense. Blind iteration over 1:N does NOT recover the full universe – most numeric IDs in that range have no data. The default ids = NULL therefore uses the active universe from cg_list(), not a numeric range. To recover delisted coins you must supply the IDs explicitly (e.g., the union of accumulated cg_list() snapshots, or the historic mapping from cg_id_mapping()).

  • Slug recovery for delisted coins is not generally available on the free tier. Active coins get their slug/name joined back in from cg_list(); rows whose numeric ID is no longer in the active universe come back with slug = NA and name = NA. Use the id column as the join key in downstream code. For a one-shot complete recovery of the full historic universe see vignette("coingecko-pro-backfill").

Examples

if (FALSE) { # \dontrun{
# Scan first 200 numeric IDs (will include both active and delisted coins)
h <- cg_history_by_id(ids = 1:200, what = c("price", "market_cap"))

# Full sweep -- survivorship-bias-free price history of the entire
# CoinGecko universe. Slow (10+ hours). Run via cronjob package.
h_all <- cg_history_by_id()
} # }