Closed
Description
For queries for which the purpose is to create/replace tables: CREATE OR REPLACE TABLE
, if the table is huge, it will eat up a lot of memory currently using read_gbq
.
And it turns out that since pandas-gbq==0.12.0
, a new argument max_results
was implemented to limit the number of rows in result dataframe for e.g. setting it to 0
for DDL queries, which will help save memory.
So it will be nice for pandas to add it for read_gbq
as pandas-gbq
does.