First I believe you mean on the Dataset API rather than the dataframe API.
You can easily add the partition index as a new column to your dataframe using
spark_partition_id()
Then a normal mapPartitions should work fine (i.e. you should create the
appropriate case class which includes the partition id and then do
mapPartitions).
Thanks,
Assaf.
From: Lalwani, Jayesh [mailto:jayesh.lalw...@capitalone.com]
Sent: Thursday, August 03, 2017 5:20 PM
To: user@spark.apache.org
Subject: mapPartitioningWithIndex in Dataframe
Are there any plans to add mapPartitioningWithIndex in the Dataframe API? Or is
there any way to implement my own mapPartitionWithIndex for a Dataframe?
I am implementing something which is logically similar to the randomSplit
function. In 2.1, randomSplit internally does df.mapPartitionWithIndex and
assigns a different seed for every partition by adding the partition’s index to
the seed. I want to get a partition specific seed too.
The problem is rdd.mapPartitionWithIndex doesn’t work in streaming.
df.mapPartition works, but I don’t get index.
Is there a way to extend Spark to add mapPartitionWithIndex at the Dataframe
level ?
I was digging into the 2.2 code a bit and it looks like in 2.2, all the
Dataframe apis have been changed to be based around SparkStrategy. I couldn’t
figure out how I can add my own custom strategy. Is there any documentation
around this? If it makes sense to add this to Spark, I would be excited to make
a contribution.
________________________________
The information contained in this e-mail is confidential and/or proprietary to
Capital One and/or its affiliates and may only be used solely in performance of
work or services for Capital One. The information transmitted herewith is
intended only for use by the individual or entity to which it is addressed. If
the reader of this message is not the intended recipient, you are hereby
notified that any review, retransmission, dissemination, distribution, copying
or other use of, or taking of any action in reliance upon this information is
strictly prohibited. If you have received this communication in error, please
contact the sender and delete the material from your computer.
