write_xpt() assigns format "$" and informat "$" by default

[jeblucas]

write_xpt() seems to apply a currency ($) format/informat designation on exported columns (at least in version = 5 which is what I need for FDA filings).

Attaching a screenshot of the output as seen in SAS Universal Viewer with issue labeled.

library(tibble)
library(haven)
try <- tibble::tibble("STUDYID" = as.character(20011234), USUBJID = as.character("20011234-5678"))
glimpse(try)
#> Observations: 1
#> Variables: 2
#> $ STUDYID <chr> "20011234"
#> $ USUBJID <chr> "20011234-5678"
write_xpt(try, file.path(getwd(),"pp.xpt"), version = 5)

[hadley]

Slightly simpler reprex:

library(haven)
path <- tempfile()
df <- tibble::tibble(x = 20011234, y = "20011234-5678")
write_xpt(df, path, version = 5, name = "test")

^{Created on 2019-11-06 by the reprex package (v0.3.0)}

@evanmiller any ideas what the problem might be here? I am calling readstat_variable_set_format(var, NULL) for each variable.

[evanmiller]

$ is a generic character format and does not indicate currency

https://support.sas.com/documentation/cdl/en/leforinforref/64790/HTML/default/viewer.htm#n134ahpcz8murvn1x7went6p2czo.htm

[hadley]

Thanks for the clarification!

[lock]

This old issue has been automatically locked. If you believe you have found a related problem, please file a new issue (with reprex) and link to this issue. https://reprex.tidyverse.org/