Got it.
Only NamedExpression have exprId, we have to make new Attribute here.
private[this] val computedSchema = computedAggregates.map(_.resultAttribute)
--
View this message in context:
http://apache-spark-developers-list.1001551.n3.nabble.com/SparkSQL-Why-this-AttributeReference-exprId-i
https://github.com/apache/spark/blob/master/sql/core/src/main/scala/org/apache/spark/sql/execution/Aggregate.scala#L85
I can't understand this code, it seems to be a "bug", but group by of
SparkSQL just works fine.
with code below, some expressions are mapping to AttributeReferences, then
"bindRe
My fork is (still) even with the master branch, so there is nothing to
update. Talking about this here: https://github.com/apache/spark/pull/3405
If anyone knows how to configure the build process on handling
warnings/errors from javadoc, that would be very helpful.
--
View this message in cont
Thanks Patrick,
You raise a good point - for this to be useful it's imperative that it is
updated with new versions of spark.
My thought with putting it on the wiki was that it's lower friction for
community members to edit, but it likely won't have the same level of
quality control as the existi
The vote passes unanimously with 4 binding +1 votes, 5 non-binding +1
votes, and no +0 or -1 votes. The final release will be posted in the next
48 hours. Thanks to everyone who voted.
-Andrew
+1:
Andrew Or*
Xiangrui Meng*
Krishna Sankar
Matei Zaharia*
Sean Owen
Anant Asthana
Marcelo Vanzin
Patr
+1 Release this package as Apache Spark 1.1.1
On 20 Nov 2014 04:22, "Andrew Or" wrote:
> I will start with a +1
>
> 2014-11-19 14:51 GMT-08:00 Andrew Or :
>
> > Please vote on releasing the following candidate as Apache Spark version
> 1
> > .1.1.
> >
> > This release fixes a number of bugs in Sp
Actually +1 from me...
This is a recommendAll feature we are testing which is really compute
intensive...
For ranking metric calculation I was trying to run through the Netflix
matrix and generate a ranked list of recommendation for all 17K products
and perhaps it needs more compute than what is
All, I just finished the SPARK-3182 feature and, for me, its raised a larger
question of how to ensure patches that are awaiting review get noted / tagged
upstream. Since I don’t have access writes to assign the above issue to myself
I can’t tag it as “In Progress” like Matei mentioned so, at th
MapOutputTrackerMasterActor sends the `mapOutputStatuses` to a buffer at
first. The messages in this buffer will be sent by some background threads.
In these threads, they will check if there are already too many messages
sent to Akka. If so, they will wait until there is enough memory.
I put a co
