I do hard code this in view. That not very efficient especially with CTAS as it
needs a 'duplicates' code : 1 from view + 1 for the parquet file (the metadata
is not handle when reading parquet file)There is an ODBC driver option:
CastAnyToVarchar=true. Not sure that can help you.Boris
Le Mardi 20 octobre 2015 19h55, Stefán Baxter <ste...@activitystream.com>
a écrit :
Hi,
What are people here doing to supply meta- and summary data for their
JSON/Parquet tables to BI tools.
Are there any best practices or does it have to be specified (hard coded)
in views.
What is being done to find low cardinality fields and their values etc.?
Regards,
-Stefán
