Github user alexbaretta commented on the pull request:
https://github.com/apache/spark/pull/3957#issuecomment-69476507
@marmbrus
Michael,
I did not realize the title could be changed. Thanks for pointing it out.
Regarding the use case: I have large, unwieldy CSV files from a data
vendor. Each comes with a machine readable schema file. In my code I parse the
schema files and construct StructType objects describing the rows of the CSV
file. Correspondingly, I create a SchemaRDD from the CSV files by parsing them
according to the schema. I then want to save these SchemaRDDs to parquet for
faster processing. The createParquetFile method allows me to create an empty
parquet directory in HDFS which I register as SparkSQL table. Finally, I call
the inserInto method on my large SchemaRDDs derived from my CSV files to save
them to the parquet files. Works quite well actually, speeding up the
computation of univariate statistics from tens of minutes on a 256 core cluster
down to less than ten seconds.
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
