I wish we can support randn function.
The reason is when i am trying to generate some sample data, it is pretty useful.
However currently it is falling back to CPU:
! <Randn> randn(xxx) cannot run on GPU because GPU does not currently support the operator class org.apache.spark.sql.catalyst.expressions.Randn
I wish we can support
randn
function. The reason is when i am trying to generate some sample data, it is pretty useful. However currently it is falling back to CPU: