Closed James-G-Hill closed 2 years ago
Thanks and agreed! Arguably col_types
should be set explicitly rather than allowing readr
to guess? At the time of writing that argument to suppress those messages didn't exist in readr
. A PR would be much appreciated!
Note that meanwhile, the user can simply toggle this behavior off as described in readr
documentation by setting
options(readr.show_col_types = FALSE)
Yeah, I'll be happy to submit a PR when I get time; it's not urgent of course. If col_types
should be set, you are suggesting the user should be able to pass the readr
col_types
parameter through to rdf_query
and then through to readr
?
Thanks. Yeah, right, ideally the user would need to be able to pass the desired col_types
to rdf_query
, since of course the return types may depend on the query. So ideally a PR would a col_types
argument to rdf_query
for that purpose. Also, if col_types
is not given by the user, (i.e. the default), then it should either suppress the warning while allowing readr to guess (as in the current behavior), but I think it would be better just to default to character
type, since it is possible for queries to result in 'mixed type' responses and string encoding is probably the safest choice to avoid accidental coercion to NAs.
I think the solution provided is good enough; not sure the col_types
is necessary; better to just remove it then add in later if somebody needs it.
When calling
rdf_query
there is a call toreadr::read_csv
. Each time this is called there is a redundant output to the console showing the data types of the data in the loaded file (always property, value, source).There should be an option to turn this off when running
rdf_query
by passing through a boolean value to thereadr::read_csv
parameter:show_col_types
.