You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Currently, NA values in a data.frame written to AnVIL with avtable_import results in the string "NA" appearing in the AnVIL data table, regardless of data type. Reading the same data table back into R with avtable results in NA for numeric columns, but the character string "NA" for character columns.
It would be helpful to control how missing values are read and written, as in the na argument to read_delim and write_delim in the readr package.
The text was updated successfully, but these errors were encountered:
## round-trip "" and NA_character_ from R to AnVIL and back
tbl0 <- tibble(
issue75 = letters[1:4],
x = c("a", "", NA, "d"),
i = c(1:2, NA, 3L),
r = as.numeric(i),
)
avtable_import(tbl0)
avtable("issue75", na = "NA")
## preserve "", NA_character_, and "NA" (sodium? North America?)
tbl0 <- tibble(
issue75 = letters[1:4],
x = c("a", "", NA, "NA"),
i = c(1:2, NA, 3L),
r = as.numeric(i),
)
avtable_import(tbl0, na = "__MISSING_VALUE__")
avtable("issue75", na = "__MISSING_VALUE__")
This follows the {readr} defaults, and change the default behavior of AnVIL (which wasn't very helpful anyway)
Currently,
NA
values in a data.frame written to AnVIL withavtable_import
results in the string "NA" appearing in the AnVIL data table, regardless of data type. Reading the same data table back into R withavtable
results inNA
for numeric columns, but the character string"NA"
for character columns.It would be helpful to control how missing values are read and written, as in the
na
argument toread_delim
andwrite_delim
in thereadr
package.The text was updated successfully, but these errors were encountered: