I fixed this issue by following commit:
https://github.com/wso2/carbon-ml/commit/d94570ed5bab36af60606d06233f1d3eecf58558#diff-7bee5ca92183e001202ee9d8b80918ca
On Wed, Aug 26, 2015 at 9:23 AM, Nirmal Fernando wrote:
> Shall we try adding an all-in-one parquet jar? Can you
Hi Nirmal,
Not actually. I have added parquet-common, parquet-encoding,
parquet-column, parquet-hadoop as mentioned at
https://github.com/Parquet/parquet-mr
https://www.google.com/url?q=https%3A%2F%2Fgithub.com%2FParquet%2Fparquet-mrsa=Dsntz=1usg=AFQjCNEHFd3g0eqXClZCkN35QgFYl0FVww
. But still get
So it's coming from 2 bundles? Can we get rid of 1 and see?
On Wed, Aug 26, 2015 at 8:40 AM, Madawa Soysa madawa...@cse.mrt.ac.lk
wrote:
Gives following output.
parquet.hadoop.metadata; version=0.0.0parquet_common_1.6.0_1.0.0 [321]
parquet.hadoop.metadata;
Madawa, hope the issue is resolved now after the instructions given offline.
On Mon, Aug 24, 2015 at 8:34 PM, Madawa Soysa madawa...@cse.mrt.ac.lk
wrote:
I have tested the model.save() with a simple Java program. It works fine.
I have noticed that scala-library:2.11.6 is a dependency of
Gives following output.
parquet.hadoop.metadata; version=0.0.0parquet_common_1.6.0_1.0.0 [321]
parquet.hadoop.metadata; version=0.0.0parquet_hadoop_1.6.0_1.0.0 [323]
On 26 August 2015 at 08:35, Nirmal Fernando nir...@wso2.com wrote:
You added them to repository/components/lib right? Can you
Both bundles are required. I tried removing one at a time and checked.
On 26 August 2015 at 08:43, Nirmal Fernando nir...@wso2.com wrote:
So it's coming from 2 bundles? Can we get rid of 1 and see?
On Wed, Aug 26, 2015 at 8:40 AM, Madawa Soysa madawa...@cse.mrt.ac.lk
wrote:
Gives following
Shall we try adding an all-in-one parquet jar? Can you check whether
there's one?
http://mvnrepository.com/artifact/com.twitter/parquet/1.6.0
On Wed, Aug 26, 2015 at 8:55 AM, Madawa Soysa madawa...@cse.mrt.ac.lk
wrote:
Both bundles are required. I tried removing one at a time and checked.
On
You added them to repository/components/lib right? Can you start the server
in OSGi mode (./wso2server.sh -DosgiConsole ) and issue following command
in OSGi console;
osgi p parquet.hadoop.metadata
On Wed, Aug 26, 2015 at 8:32 AM, Madawa Soysa madawa...@cse.mrt.ac.lk
wrote:
Hi Nirmal,
Not
I have tested the model.save() with a simple Java program. It works fine.
I have noticed that scala-library:2.11.6 is a dependency of
spark:spark-core_2.11:1.4.1 [1]
http://search.maven.org/#artifactdetails%7Corg.apache.spark%7Cspark-core_2.11%7C1.4.1%7Cjar
In ML scala version is explicitly
Madawa, seems this is a Spark issue :-( can you try a simple Java program
and see whether model.save() works?
On Sat, Aug 22, 2015 at 8:19 AM, Madawa Soysa madawa...@cse.mrt.ac.lk
wrote:
Hi Nirmal,
outPath is correct and the directory gets created, but the process becomes
idle after that.
Hi Nirmal,
outPath is correct and the directory gets created, but the process becomes
idle after that. Attached the only part that was written to a file.
Also the method doesn't throw an exception as well.
On 21 August 2015 at 21:31, Nirmal Fernando nir...@wso2.com wrote:
Hi Madawa,
Hi Madawa,
According to Spark API [1], outPath shouldn't be exist.
[1]
https://github.com/apache/spark/blob/master/mllib/src/main/scala/org/apache/spark/mllib/recommendation/MatrixFactorizationModel.scala#L200
On Fri, Aug 21, 2015 at 1:59 PM, Niranda Perera nira...@wso2.com wrote:
I don't
I don't think it's correct. Scala version is 2.10.4 even in the mvn repo
On Fri, Aug 21, 2015, 13:46 Madawa Soysa madawa...@cse.mrt.ac.lk wrote:
Also I asked this question in StackOverflow[1]
Also I asked this question in StackOverflow[1]
http://stackoverflow.com/questions/32048618/how-to-serialize-apache-sparks-matrixfactorizationmodel-in-java
and there they have mentioned a version incompatibility between Scala and
Spark versions
[1] -
Yes path is valid, I explicitly set the path here from the
MLModelHandler persistModel method.
On 21 August 2015 at 10:26, Nirmal Fernando nir...@wso2.com wrote:
On Thu, Aug 20, 2015 at 9:21 PM, Madawa Soysa madawa...@cse.mrt.ac.lk
wrote:
Hi All,
There an issue with serializing Spark's
On Thu, Aug 20, 2015 at 9:21 PM, Madawa Soysa madawa...@cse.mrt.ac.lk
wrote:
Hi All,
There an issue with serializing Spark's MatrixFactorizationModel object.
The object contains a huge RDD and as I have read in many blogs, this model
cannot be serialized as a java object. Therefore when
Hi All,
There an issue with serializing Spark's MatrixFactorizationModel object.
The object contains a huge RDD and as I have read in many blogs, this model
cannot be serialized as a java object. Therefore when retrieving the model
I get the same exception as above;
*Caused by:
Yes I was able to resolve the issue by removing RDD fields from the
SummaryModel object as @Mano pointed out. Still I have the same exception
when retrieving the model. Trying to fix that issue.
On 14 August 2015 at 10:43, Nirmal Fernando nir...@wso2.com wrote:
Thanks Niranda, this doc is
Thanks Niranda, this doc is useful.
On Fri, Aug 14, 2015 at 10:36 AM, Niranda Perera nira...@wso2.com wrote:
From what I know, OneToOneDependancy come into play when spark tries to
create the RDD dependency tree.
Just thought of sharing that. this would be a good resource [1] :-)
[1]
Were you able to resolve the issue?
On Thu, Aug 13, 2015 at 8:59 AM, Nirmal Fernando nir...@wso2.com wrote:
@Mano can you help Madawa?
On Thu, Aug 13, 2015 at 8:56 AM, Madawa Soysa madawa...@cse.mrt.ac.lk
wrote:
No that's not something I use.
On 13 August 2015 at 00:09, Nirmal Fernando
From what I know, OneToOneDependancy come into play when spark tries to
create the RDD dependency tree.
Just thought of sharing that. this would be a good resource [1] :-)
[1]
https://databricks-training.s3.amazonaws.com/slides/advanced-spark-training.pdf
On Thu, Aug 13, 2015 at 12:09 AM,
@Mano can you help Madawa?
On Thu, Aug 13, 2015 at 8:56 AM, Madawa Soysa madawa...@cse.mrt.ac.lk
wrote:
No that's not something I use.
On 13 August 2015 at 00:09, Nirmal Fernando nir...@wso2.com wrote:
What is *org.apache.spark.OneToOneDependency ? Is it something you use?*
On Wed, Aug
No that's not something I use.
On 13 August 2015 at 00:09, Nirmal Fernando nir...@wso2.com wrote:
What is *org.apache.spark.OneToOneDependency ? Is it something you use?*
On Wed, Aug 12, 2015 at 11:30 PM, Madawa Soysa madawa...@cse.mrt.ac.lk
wrote:
Hi,
I created a model summary in order
What is *org.apache.spark.OneToOneDependency ? Is it something you use?*
On Wed, Aug 12, 2015 at 11:30 PM, Madawa Soysa madawa...@cse.mrt.ac.lk
wrote:
Hi,
I created a model summary in order to show the model data in the
analysis.jag page.
But when refreshing the page after building the
Hi,
I created a model summary in order to show the model data in the
analysis.jag page.
But when refreshing the page after building the model I get the following
error.
org.wso2.carbon.ml.core.exceptions.MLAnalysisHandlerException: An error
has occurred while extracting all the models of
25 matches
Mail list logo