Question about the different number of records obtained when using get_lsratio and get_quote

[andreltr]

Hello,

I'm getting a very different number of records result when using get_quote and get_lsratio. See below, please.

tail(
BTC <- get_quote(
ticker = 'BTCUSDT',
source = 'binance',
interval = '1d',
from = Sys.Date() %m-% months(48),
to = Sys.Date()
)
)
tail(
BTC_LS <- get_lsratio(
ticker = 'BTCUSDT',
source = 'binance',
interval = '1d',
from = Sys.Date() %m-% months(48),
to = Sys.Date()
)
)

R> nrow(BTC)
[1] 1462
R> nrow(BTC_LS)
[1] 29
R>

Is this expected? I was considering creating a dataframe including the BTC_LS fields joined with the BTC dataframe but it didn't work as intended.

One thing I imagined was to add the BTC_LS records in the common index BTC records but that would leave empty records in most of the new BTC columns which were joined from the BTC_LS dataframe.

The objective is to have a BTC dataframe that includes all the information required for a more chart, such as one including the bollinger_bands(), ema(), sma(), and wma() indicators along with volume() and long-short ratio. The graph I get depicts the L-S ratio compressed with all its itens very close or on top of each other and I guess it should be spread out through the common range.

TIA,

André Luiz

[serkor1]

Hi @andreltr,

Thank you for another great question. I have moved your question to general as it highlights some important aspects of {cryptoQuotes} in general.

Your question is a good example of restrictions imposed on exchange-level as we have discussed in this post. Binance only supports get_lsratio() from the last 30 days - so the behavior is expected. But it's undocumented; I will update the wiki, and probably the documentation accordingly.

A work around is to use Bybit for this instead, below is an example,

library(cryptoQuotes)

# get long-short
# ratioes
ticker <- get_lsratio(
  ticker = "BTCUSDT",
  source = "bybit",
  from   = Sys.Date() - 750,
  to     = Sys.Date()
)

# see oldest record
head(ticker)
#>                       long  short  ls_ratio
#> 2023-02-11 01:00:00 0.5259 0.4741 1.1092596
#> 2023-02-12 01:00:00 0.5253 0.4747 1.1065936
#> 2023-02-13 01:00:00 0.5255 0.4745 1.1074816
#> 2023-02-14 01:00:00 0.5273 0.4727 1.1155067
#> 2023-02-15 01:00:00 0.5092 0.4908 1.0374898
#> 2023-02-16 01:00:00 0.4945 0.5055 0.9782394

# count records
nrow(ticker)
#> [1] 500

Please note that we have used Sys.Date() - 750 but only got 500 rows of data. This is another case of exchange imposed restrictions; Bybit only support long-short ratios from the last 500 days.

My recommendation is that you explore the different exchanges, and how they behave in the different endpoints. I will make sure to update the documentation by version 1.3.2.

^{Created on 2024-06-24 with reprex v2.1.0}