knitr::opts_chunk$set( collapse = TRUE, comment = "#>", fig.path = "man/figures/README-" )
This is a modification of the original crypto
package by jesse vent. It is entirely set up to use means from the tidyverse
and provides tibble
s with all data available via the web-api of coinmarketcap.com. It does not require an API key but in turn only provides information that is also available through the website of coinmarketcap.com.
It allows the user to retrieve
crypto_listings()
a list of all coins that were historically listed on CMC (main dataset to avoid delisting bias) according to the CMC API documentationcrypto_list()
a list of all coins that are listed as either being active, delisted or untracked according to the CMC API documentationcrypto_info()
a list of all information available for all available coins according to the CMC API documentationcrypto_history()
the most powerful function of this package that allows to download the entire available history for all coins covered by CMC according to the CMC API documentationcrypto_global_quotes()
a dataset of historical global crypto currency market metrics to the CMC API documentationfiat_list()
a mapping of all fiat currencies (plus precious metals) available via the CMC WEB APIexchange_list()
a list of all exchanges available as either being active, delisted or untracked according to the CMC API documentationexchange_info()
a list of all information available for all given exchanges according to the CMC API documentationSlight change in api output broke crypto_info()
(new additional column). Fixed.
Slight change in api output broke crypto_info()
. Fixed.
After a major change in the api structure of coinmarketcap.com, the package had to be rewritten. As a result, many functions had to be rewritten, because data was not available any more in a similar format or with similar accuracy. Unfortunately, this will potentially break many users implementations. Here is a detailed list of changes:
crypto_list()
has been modified and delivers the same data as before.exchange_list()
has been modified and delivers the same data as before.fiat_list()
has been modified and no longer delivers all available currencies and precious metals (therefore only USD and Bitcoin are available any more).crypto_listings()
needed to be modified, as multiple base currencies are not available any more. Also some of the fields downloaded from CMC might have changed. It still retrieves the latest listings, the new listings as well as historical listings. The fields returned have somewhat slightly changed. Also, no sorting is available any more, so if you want to download the top x CCs by market cap, you have to download all CCs and then sort them in R.crypto_info()
has been modified, as the data structure has changed. The fields returned have somewhat slightly changed.crypto_history()
has been modified. It still retrieves all the OHLC history of all the coins, but is slower due to an increased number of necessary api calls. The number of available intervals is strongly limited, but hourly and daily data is still available. Currently only USD and BTC are available as quote currencies through this library.crypto_global_quotes()
has been modified. It still produces a clear picture of the global market, but the data structure has somewhat slightly changed.Since version 1.4.6 I have added the possibility to "sort" the historical crypto_listings()
in _asc_ending or _desc_ending order ("sort_dir") to allow for the possibility to download only the top x crypto currencies using "limit" based on the requested sort (not available for "new" sorting). Also corrected some problems when sourcing lists that now do not have the "last_historical_data" field available any more.
Since version 1.4.5 I have added a new function crypto_global_quotes()
which retrieves global aggregate market statistics for CMC. There also were some bugs fixed.
Since version 1.4.4 a new function crypto_listings()
was introduced that retrieves new/latest/historical listings and listing information at CMC. Additionally some aspects of the other functions have been reworked. We noticed that finalWait = TRUE
does not seem to be necessary at the moment, as well as sleep
can be set to '0' seconds. If you experience strange behavior this might be due to the the api sending back strange (old) results. In this case let sleep = 60
(the default) and finalWait = TRUE
(the default).
Since version 1.4.0 the package has been reworked to retrieve as many assets as possible with one api call, as there is a new "feature" introduced by CMC to send back the initially requested data for each api call within 60 seconds. So one needs to wait 60s before calling the api again. Additionally, since version v1.4.3 the package allows for a data interval
larger than daily (e.g. '2d' or '7d' or 'weekly')
You can install crypto2
from CRAN with
install.packages("crypto2")
or directly from github with:
# install.packages("devtools") devtools::install_github("sstoeckl/crypto2")
The package provides API free and efficient access to all information from https://coinmarketcap.com that is also available through their website. It uses a variety of modification and web-scraping tools from the tidyverse
(especially purrr
).
As this provides access not only to active coins but also to those that have now been delisted and also those that are categorized as untracked, including historical pricing information, this package provides a valid basis for any Asset Pricing Studies based on crypto currencies that require survivorship-bias-free information. In addition to that, the package maintainer is currently working on also providing delisting returns (similarly to CRSP for stocks) to also eliminate the delisting bias.
First we load the crypto2
-package and download the set of active coins from https://coinmarketcap.com (additionally one could load delisted coins with only_Active=FALSE
as well as untracked coins with add_untracked=TRUE
).
library(crypto2) library(dplyr) # List all active coins coins <- crypto_list(only_active=TRUE)
Next we download information on the first three coins from that list.
# retrieve information for all (the first 3) of those coins coin_info <- crypto_info(coins, limit=3, finalWait=FALSE) # and give the first two lines of information per coin coin_info
In a next step we show the logos of the three coins as provided by https://coinmarketcap.com.
coin_info$logo %>% knitr::include_graphics(.)
In addition we show tags provided by https://coinmarketcap.com.
coin_info %>% select(slug,tags) %>% tidyr::unnest(tags) %>% group_by(slug) %>% slice(1,n())
Additionally: Here are some urls pertaining to these coins as provided by https://coinmarketcap.com.
coin_info %>% pull(urls) %>% .[[1]] |> unlist()
In a next step we download time series data for these coins.
# retrieve historical data for all (the first 3) of them coin_hist <- crypto_history(coins, limit=3, start_date="20210101", end_date="20210105", finalWait=FALSE) # and give the first two times of information per coin coin_hist %>% group_by(slug) %>% slice(1:2)
Similarly, we could download data on an hourly basis.
# retrieve historical data for all (the first 3) of them coin_hist_m <- crypto_history(coins, limit=3, start_date="20210101", end_date="20210102", interval ="1h", finalWait=FALSE) # and give the first two times of information per coin coin_hist_m %>% group_by(slug) %>% slice(1:2)
Alternatively, we could determine the price of these coins in other currencies. A list of such currencies is available as fiat_list()
fiats <- fiat_list() fiats
So we download the time series again depicting prices in terms of Bitcoin and Euro (note that multiple currencies can be given to convert
, separated by ",").
# retrieve historical data for all (the first 3) of them coin_hist2 <- crypto_history(coins, convert="USD", limit=3, start_date="20210101", end_date="20210105", finalWait=FALSE) # and give the first two times of information per coin coin_hist2 %>% group_by(slug,ref_cur_name) %>% slice(1:2)
As a new features in version 1.4.4. we introduced the possibility to download historical listings and listing information (add quote = TRUE
).
latest_listings <- crypto_listings(which="latest", limit=10, quote=TRUE, finalWait=FALSE) latest_listings
An additional feature that was added in version 1.4.5 retrieves global aggregate market statistics for CMC.
all_quotes <- crypto_global_quotes(which="historical", quote=TRUE) all_quotes
We can use those quotes to plot information on the aggregate market capitalization:
all_quotes %>% select(timestamp, USD_total_market_cap, USD_altcoin_market_cap) %>% tidyr::pivot_longer(cols = 2:3, names_to = "Market Cap", values_to = "bn. USD") %>% tidyr::separate(`Market Cap`,into = c("Currency","Type","Market","Cap")) %>% dplyr::mutate(`bn. USD`=`bn. USD`/1000000000) %>% ggplot2::ggplot(ggplot2::aes(x=timestamp,y=`bn. USD`,color=Type)) + ggplot2::geom_line() + ggplot2::labs(title="Market capitalization in bn USD", subtitle="CoinMarketCap.com")
Last and least, one can get information on exchanges. For this download a list of active/inactive/untracked exchanges using exchange_list()
:
exchanges <- exchange_list(only_active=TRUE) exchanges
and then download information on "binance" and "bittrex":
ex_info <- exchange_info(exchanges %>% filter(slug %in% c('binance','kraken')), finalWait=FALSE) ex_info
Then we can access information on the fee structure,
ex_info %>% select(contains("fee"))
or the fiat currencies allowed:
ex_info %>% select(slug,fiats) %>% tidyr::unnest(fiats)
This project is licensed under the MIT License - see the
crypto
-package that inspired this package.Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.