Converting XTS and ZOO objects

## load library
library(cryptoQuotes)

If you want to use the simplicity of tidverse or the power of data.table, the xts-object can be easily converted.

However, its important to maintain data integrity, especially, if the date and timezone is important for you.

Converting xts and zoo to tibble

Converting to tibble requires a few steps to achieve the same data structure as the xts,

# 1) load pipe
library(magrittr)

# 2) convert to tibble
# using as_tibble
tbl <- tibble::as_tibble(
  x = cbind(
    Index = zoo::index(ATOM),
    zoo::coredata(ATOM)
  )
) %>% dplyr::mutate(
  Index = lubridate::as_datetime(
    Index
  )
)

# 3) head data
head(tbl, 3)
#> # A tibble: 3 × 6
#>   Index                open  high   low close volume
#>   <dttm>              <dbl> <dbl> <dbl> <dbl>  <dbl>
#> 1 2023-12-29 23:00:00  10.9  10.9  10.9  10.9  3260.
#> 2 2023-12-29 23:15:00  10.9  11.0  10.9  11.0  1863.
#> 3 2023-12-29 23:30:00  11.0  11.0  10.9  11.0  1861.

Converting xts and zoo to data.table

Converting to data.table is straightforward as as.data.table() handles everything under the hood,

# 1) convert to data.table
# using as.data.table
DT <- data.table::as.data.table(
  ATOM
) 

# 2) set column name to upper
colnames(DT)[1] <- 'Index'

# 3) head data
head(DT, 3)
#>                  Index    open    high     low   close   volume
#>                 <POSc>   <num>   <num>   <num>   <num>    <num>
#> 1: 2023-12-30 00:00:00 10.8778 10.9133 10.8775 10.9089 3259.760
#> 2: 2023-12-30 00:15:00 10.9089 10.9530 10.9008 10.9512 1862.697
#> 3: 2023-12-30 00:30:00 10.9512 10.9884 10.9373 10.9832 1861.493

Checking data integrity

Checking date integrity

It is important that the date.time has not been converted to a different timezone in the process, without explicitly coding it as such,

# 1) store date.time objects
time_objects <- list(
  tbl = tbl$Index,
  DT  = DT$Index
)

# 2) check if they are all equal
all(
  sapply(
    time_objects,
    function(x) {
      
      setequal(
        x = x,
        y = zoo::index(ATOM)
        )
      
    }
  )
)
#> [1] TRUE

Checking OHLCV values

It goes without saying that R-functions wouldn’t tamper with the order of the data during conversion without a warning in the documentation, but nonetheless for the sake of argument, we will check the OHLCV values,

# 1) store price objects
# Open price here
open_price <- list(
  tbl = tbl$Open,
  DT  = DT$Open
)
#> Warning: Unknown or uninitialised column: `Open`.


# 2) check if they are all equal
all(
  sapply(
    open_price,
    function(x) {
      
      setequal(
        x = x,
        y = ATOM$Open
        )
      
    }
  )
)
#> [1] TRUE

Why even convert?

Even though numerical operations on xts-objects are lightning fast (12% faster than data.table), it comes with a cost; it doesn’t support factors or characters.