Skip to contents

Convert a numpy 64-bit integer array to R

Source: R/int64.R
pyids2bit64.Rd

Parses a numpy array (or Python list/int) of 64-bit integer ids into a bit64::integer64 vector or character, round-tripping through raw bytes so no precision is lost.

Usage

pyids2bit64(x, as_character = TRUE)

Arguments

x

A numpy array, Python list or Python int of 64-bit integers.

as_character

If TRUE (the default) return a character vector; otherwise a bit64::integer64 vector.

Value

A character or bit64::integer64 vector.

Details

Only int64 and uint64 numpy dtypes are accepted; a uint64 array is checked for values that would overflow signed int64 and errors if any is found. The bytes are written to a temporary file and read back as doubles whose class is then set to integer64, which is exact for all 64-bit ids.

See also

rids2pyint() for the reverse direction.

Examples

if (FALSE) { # \dontrun{
np <- reticulate::import("numpy", convert = FALSE)
pyids2bit64(np$array(c("123", "456"), dtype = "uint64"))
} # }