powered by
Coalesces a Spark DataFrame
sdf_coalesce(x, partitions)
A spark_connection, ml_pipeline, or a tbl_spark.
spark_connection
ml_pipeline
tbl_spark
number of partitions