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
Is your feature request related to a problem? Please describe.
Reading floating-point numbers from CSV and JSON is a common operation and we would like it to be enabled by default. It was disabled by default in 22.04 due to cuDF having different precision to Java and therefore results were not the same as Spark. Now that this issue is resolved in rapidsai/cudf#10622 we should be able to enable these types by default.
Describe the solution you'd like
Enable reading float and double from CSV and JSON by default and remove the associated configuration options.
Describe alternatives you've considered
None
Additional context
None
The text was updated successfully, but these errors were encountered:
Is your feature request related to a problem? Please describe.
Reading floating-point numbers from CSV and JSON is a common operation and we would like it to be enabled by default. It was disabled by default in 22.04 due to cuDF having different precision to Java and therefore results were not the same as Spark. Now that this issue is resolved in rapidsai/cudf#10622 we should be able to enable these types by default.
Describe the solution you'd like
Enable reading float and double from CSV and JSON by default and remove the associated configuration options.
Describe alternatives you've considered
None
Additional context
None
The text was updated successfully, but these errors were encountered: