Have you exposed the Attributes on the Workspace?

“SELECT *” gives no clues to SQLExecutor as to what Workspace Attribute Names to create for generated Workspace Features.  This will need to be set in the Attributes to Expose Parameter, or in a downstream AttributeExposer

​
 

If you however view the results in Data Inspector you will see on any individual Feature in the Feature Information any unexposed Attribute values.

​@oliver.morris : I’m getting the same: no output when using SUMMARIZE, not even unexposed output.

And its sister statement DESCRIBE even gives an error:

​@oliver.morris : There is a very simple workaround: wrap the SUMMARIZE or DESCRIBE query in another SELECT * FROM clause:

SELECT * FROM (SUMMARIZE SELECT * FROM read_csv('https://oedi-data-lake.s3.amazonaws.com/pvdaq/csv/systems.csv') );

And in DuckDB’s SQL dialect you don’t explicitly need the “SELECT *” or the read_csv() function, so this could be abbreviated to:

FROM ( SUMMARIZE FROM 'https://oedi-data-lake.s3.amazonaws.com/pvdaq/csv/systems.csv' );

​@oliver.morris : There is a very simple workaround: wrap the SUMMARIZE or DESCRIBE query in another SELECT * FROM clause:

SELECT * FROM (SUMMARIZE SELECT * FROM read_csv('https://oedi-data-lake.s3.amazonaws.com/pvdaq/csv/systems.csv') );

And in DuckDB’s SQL dialect you don’t explicitly need the “SELECT *” or the read_csv() function, so this could be abbreviated to:

FROM ( SUMMARIZE FROM 'https://oedi-data-lake.s3.amazonaws.com/pvdaq/csv/systems.csv' );

​@arnold_bijlsma nice solution, very cool - I think this means I can build a little hub transformer to summarise a whole set of file types super quick!