his between public and private lists is
confusing, and not very ASF friendly.
John
On Sat, Jul 15, 2017 at 5:39 PM Andrew Palumbo wrote:
> + dev@flink (if they allow this through moderation.. I'm not subscribed
> to their list anymore; If not I believe that there are enough Flink PMCs
+ dev@flink (if they allow this through moderation.. I'm not subscribed to
their list anymore; If not I believe that there are enough Flink PMCs on
private@mahout to answer
We've stopped shipping the flink binaries as of Mahout 0.13.0:
https://mvnrepository.com/artifact/org.apache.mahout/maho
Greg,
We ran into this Issue when implementing the Mahout bindings for Flink [1]. It
ended up being the major bottleneck for Mahout on Flink, and makes iterative
algorithms basically unreasonable. While it is understook that that Flink's
Delta-iterations are intended for use when iterating ov
t; compute equally sized partitions.
>
> The DataDistribution interface has a method to return the bucket boundaries
> among which the partitions are divided.
>
> Please let me know if I didn't get the requirements right or if you have
> further questions.
>
> Best, Fab
Also I shoud say that i doo need the data in the partitions to remain in the
same order. Ie. partition 3 can not go into partiton 0 if the partition 0 is
empty and there are 2 of 4 partitions being used.
From: Andrew Palumbo
Sent: Saturday, May 14
ers,
Till
On Mon, Apr 25, 2016 at 8:18 PM, Andrew Palumbo wrote:
> sorry - just noticed below should read:
>
> val rowsA = (0 until inCoreA.nrow).map(i => (i, inCoreA(i, ::)))
> drmA = env.fromCollection(rowsA).partitionByRange(0)
>
> val rowsB = (0 until
ange(0)
also:
The Blockified representation is a `DataSet[(Array(K), Matrix)]`.
Thanks
From: Andrew Palumbo
Sent: Monday, April 25, 2016 1:58 PM
To: dev@flink.apache.org
Subject: Re: Partition problem
Thank you Fabian and Till for answering,
I think t
ock
> index. That's because the distribution of the individual rows in the
> cluster is not necessarily the same between two matrices.
>
> Cheers,
> Till
>
> On Mon, Apr 25, 2016 at 1:40 AM, Andrew Palumbo
> wrote:
>
> > Hi All,
> >
> >
> > I'
Hi All,
I've run into a problem with empty partitions when the number of elements in a
DataSet is less than the Degree of Parallelism. I've created a gist here to
describe it:
https://gist.github.com/andrewpalumbo/1768dac6d2f5fdf963abacabd859aaf3
I have two 2x2 matrices, Matrix A and Matri
Andrew Palumbo created FLINK-3762:
-
Summary: Kryo StackOverflowError due to disabled Kryo Reference
tracking
Key: FLINK-3762
URL: https://issues.apache.org/jira/browse/FLINK-3762
Project: Flink
Do you want me to open a jira/pr for this?
Original message
From: Stephan Ewen
Date: 04/13/2016 5:16 AM (GMT-05:00)
To: dev@flink.apache.org
Subject: Re: Kryo StackOverflowError
+1 to add this to 1.0.2
On Wed, Apr 13, 2016 at 1:57 AM, Andrew Palumbo wrote:
>
&
then it tried to serialize class A. To
> > > serialize class A it tried to serialize the method createObjects. Or
> > > something like that, I do not remember the details. This caused the
> > > recursion.
> > >
> > > BR,
> > > Hilmi
> > >
> &
e not the whole graph is recursively
processed.
Can you paste the entire Stack Trace (for example to a gist)?
Greetings,
Stephan
On Sun, Apr 10, 2016 at 4:42 AM, Andrew Palumbo wrote:
> Hi all,
>
>
> I am working on a matrix multiplication operation for Mahout Flink
> Bindings
Hi all,
I am working on a matrix multiplication operation for Mahout Flink Bindings
that uses quite a few chained Flink Dataset operations,
When testing, I am getting the following error:
{...}
04/09/2016 22:30:35CHAIN Reduce (Reduce at
org.apache.mahout.flinkbindings.blas.FlinkOpABt$.
Andrew Palumbo created FLINK-3605:
-
Summary: DataSet.output(HadoopOutput) always expects a
LongWritable as a key and throws an exception when supplied with other Writable
Key Types
Key: FLINK-3605
URL: https
The Apache Mahout PMC is pleased to announce the release of Mahout 0.11.0.
Mahout's goal is to create an environment for quickly creating machine learning
applications that scale and run on the highest performance parallel computation
engines available. Mahout comprises an interactive environmen
16 matches
Mail list logo
