sstoeckl/crypto2

Cryptocurrency Market Data

crypto2's Issues

crypto_history limit needs to be set with coin_list argument set

To start with, amazing package!

Not sure if intended but when filtering on a single coin in the coin_list argument by first using the crypto_list (within the crypto_history) function) it looks like the limit needs to be set in order for the function to work properly. It gives an error when the limit argument is kept at default (NULL) and when provided it runs.

Not setting the limit argument and keeping it to NULL (Default) causes error below.

crypto2/R/crypto_history.R

Line 77 in 3059e48

if (!is.null(limit)) coin_list <- coin_list[1:limit, ]

How to retreive historical global/aggregated market cap data?

Hi there,

I would like to retreive the historical global/aggregated marketcap over all coins as shown in the first chart here: https://coinmarketcap.com/charts/

Unfortunately, I haven't found the respective function in the package reference.

Any help would be highly appreciated!

Thank you very much in advance for a short reply.

Best,
André

Is it still free to get the historical data?

I can see in the documentation that

https://coinmarketcap.com/api/documentation/v1/#operation/getV1CryptocurrencyListingsHistorical

This endpoint is available on the following API plans - the cheapest plans don't allow to use this endpoint.

I'm not R developer, what I'd like to get are the prices of top N market-cap cryptos over a few years by day.

Can I expect crypto2 to allow me to get this data?

Faster queries for short amount of data?

Hi, thanks for making this package, this is really useful.
I was trying out the crypto_history function and it seems that no matter if you put a short timeline to cover (just the past 7 days for example), it fetches the whole data anyway and executes a filter at the end. Would it be possible to fetch only the data of interest instead to make for much faster queries?
Thanks!

Daily data

Use data.frame over tibble

Thanks for this great package. Is there any reason for using tibble over data.frame as the data format? Tibble is as slow as data.frame but more noisey, and can cause issues when converted to data.table. Would you consider using data.frame as the data format?

Create loop to do the history download in chunks of coins

Create loop to do the history download in chunks of coins. It takes very long already and needs quite some RAM and time. Can (of cousre) be done manually.

Hourly Data

Love the package. Thanks for developing this. Question: Do you have plans to integrate the "hourly" interval by chance? If not, I recommend removing it from the docs.

crypto_listings - Error in open.connection(con, "rb")

This:
> crypto_listings(which="latest", limit=10, quote=TRUE, finalWait=FALSE)
Gives:

Error in open.connection(con, "rb") : 
  cannot open the connection to 'https://web-api.coinmarketcap.com/v1/cryptocurrency/listings/latest?limit=10&aux=market_cap_by_total_supply&convert=USD&sort=cmc_rank&sort_dir=desc&start=1'
In addition: Warning message:
In open.connection(con, "rb") :
  cannot open URL 'https://web-api.coinmarketcap.com/v1/cryptocurrency/listings/latest?limit=10&aux=market_cap_by_total_supply&convert=USD&sort=cmc_rank&sort_dir=desc&start=1': HTTP status was '400 Bad Request'

sessionInfo()

> sessionInfo()
R version 4.2.1 (2022-06-23)
Platform: x86_64-apple-darwin17.0 (64-bit)
Running under: macOS Monterey 12.6

Matrix products: default
LAPACK: /Library/Frameworks/R.framework/Versions/4.2/Resources/lib/libRlapack.dylib

locale:
[1] en_US.UTF-8/en_US.UTF-8/en_US.UTF-8/C/en_US.UTF-8/en_US.UTF-8

attached base packages:
[1] stats     graphics  grDevices utils     datasets  methods   base     

other attached packages:
[1] dplyr_1.1.2      crypto2_1.4.5.99

loaded via a namespace (and not attached):
 [1] progress_1.2.2    zoo_1.8-12        tidyselect_1.2.0  remotes_2.4.2     purrr_1.0.1       lattice_0.21-8   
 [7] vctrs_0.6.3       generics_0.1.3    miniUI_0.1.1.1    htmltools_0.5.5   usethis_2.1.6     utf8_1.2.3       
[13] rlang_1.1.1       pkgbuild_1.4.0    later_1.3.1       pillar_1.9.0      urlchecker_1.0.1  glue_1.6.2       
[19] withr_2.5.0       sessioninfo_1.2.2 plyr_1.8.8        lifecycle_1.0.3   stringr_1.5.0     devtools_2.4.5   
[25] htmlwidgets_1.6.2 memoise_2.0.1     callr_3.7.3       fastmap_1.1.1     httpuv_1.6.10     ps_1.7.5         
[31] curl_5.0.0        fansi_1.0.4       Rcpp_1.0.10       xtable_1.8-4      promises_1.2.0.1  cachem_1.0.8     
[37] desc_1.4.2        pkgload_1.3.2     jsonlite_1.8.4    mime_0.12         fs_1.6.2          hms_1.1.3        
[43] digest_0.6.31     stringi_1.7.12    processx_3.8.1    shiny_1.7.4       RJSONIO_1.3-1.8   grid_4.2.1       
[49] rprojroot_2.0.3   cli_3.6.1         tools_4.2.1       magrittr_2.0.3    tibble_3.2.1      profvis_0.3.8    
[55] tidyr_1.3.0       crayon_1.5.2      pkgconfig_2.0.3   ellipsis_0.3.2    prettyunits_1.1.1 timechange_0.2.0 
[61] lubridate_1.9.2   rstudioapi_0.14   R6_2.5.1          compiler_4.2.1   

crypto_list() error

This package has always worked for me before, but today I am getting this error when I use crypto_list():

Error: Problem with mutate() input ..1.
i ..1 = dplyr::across(7:8, as.Date).
x 'origin' must be supplied

Error in if (limit == 1) { : argument is of length zero

Hello

crypto_history() returns the above error in most cases. It is particularly frustrating when it happens after a very long download time, as nothing is saved due to the error. Is there a way around this? Thanks

> data_1 <- crypto_history(start_date = "20130101", end_date = "20130201", sleep = 1)
> Scraping historical crypto data

| [32 / 32] [=====================================================] 100% in 00:32:00 ETA: 0s
Error in if (limit == 1) { : argument is of length zero
> data_1
Error: object 'data_1' not found

Self-reported tags need to get "nested"

nest self-reported tags (the coin-info-list has more entries than the coin-list due to the self-reported tags). Current workaround:
ALLINFO %>% group_by(slug) %>% nest(self_reported_tags = c(self_reported_tags))

convert tibble to data.frame

Thanks for this great R package! I used the crypto_history function to get all historical crypto prices from CoinMarketCap and got a tibble. When I convert this tibble to a data.frame, the resulted data.frame only has 11 rows. Not sure how this issue happened.

Improvements

For the next release we implement:

• a "single_id" switch for crypto_history (to not fail the entire download when one coin in id-list is missing)
• a "wait" selector to reduce waiting time between/after failed download to self-selected values (is automated at 60s as previously this was the servers default "you can load new data now" method for the free-api)

rate limiting by coinmarketcap without API key

I could only get 100 requests out before it threw HTTP error 429 for the rest of the day.

Error in interval time smaller than daily

When trying to collect historical price data of e.g., BTC for time intervals smaller than daily, I run into errors.

For example:
CMC_coin_data <- crypto_history(selected_coin, finalWait=FALSE, interval="hourly")

will give this error:
In crypto_history(selected_coin, finalWait = FALSE, interval = "hourly") :
interval was not valid, using "daily". see documentation for allowed values.

However, these work just fine:
CMC_coin_data <- crypto_history(selected_coin, finalWait=FALSE, interval="daily")
CMC_coin_data <- crypto_history(selected_coin, finalWait=FALSE, interval="monthly")

Which is strange, especially since the documentation in the crypto_history.R file states the following:

#' @param interval string Interval with which to sample data, default 'daily'. Must be one of "hourly" "daily" "weekly" #' "monthly" "yearly" "1d" "2d" "3d" "7d" "14d" "15d" "30d" "60d" "90d" "365d"

Or am I missing something here? Thank you!

Is there a way to categorize coins by consensus mechanism?

Hey there,

First of all, I just wanted to say that the package is really really good and I am a very happy user!

I was wondering whether there is a way to retrieve a list of coins (and time series) based on consensus mechanism such as "proof-of-work"?

For instance, there is a list of pow coins on CMC:
https://coinmarketcap.com/view/pow/

I have also found a category identifier in the API description on CMC called "consensus_mechanism" (see attachment).

That way crypto2 package users would be able to create coin indices based on consensus mechanism as well.

Thank you very much in advance for a short reply!

Best regards,
André