[GitHub] incubator-predictionio issue #285: PredictionIO install script doesn't work!

[rrcook]

Github user rrcook commented on the issue:

https://github.com/apache/incubator-predictionio/issues/285
  
@gdoslu 's solution did not work, @EmergentOrder 's did.
I changed the url in the quick install, substituted develop for master, it 
ran.



---
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 infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---

[GitHub] incubator-predictionio issue #274: pio train error

[steven-haddix]

Github user steven-haddix commented on the issue:

https://github.com/apache/incubator-predictionio/issues/274
  
Did you resolve this?


---
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 infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---

[GitHub] incubator-predictionio issue #285: PredictionIO install script doesn't work!

[piotremis]

Github user piotremis commented on the issue:

https://github.com/apache/incubator-predictionio/issues/285
  
There is another problem with installation script...it try to download 
`https://github.com/PredictionIO/PredictionIO/releases/download/v0.9.6/PredictionIO-0.9.6.tar.gz`
 file but it doesn't exist...
`echo "Downloading PredictionIO..."
curl -OL 
https://github.com/PredictionIO/PredictionIO/releases/download/v${PIO_VERSION}/${PIO_FILE}`


---
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 infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---

[jira] [Comment Edited] (PIO-30) Cross build for different versions of scala and spark

[Pat Ferrel (JIRA)]

[ 
https://issues.apache.org/jira/browse/PIO-30?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15449714#comment-15449714
 ] 

Pat Ferrel edited comment on PIO-30 at 8/30/16 6:18 PM:


How does Spark do their multiple version support? They have releases on 1.x and 
2.x because they are not source compatible and I think we are in the same 
situation at least for code that calls the new libs (here I only have 
experience with ES) 

They even maintain versions of the docs site for 1.x and 2.x

We might as well ask these questions now even if we don't want to do it that 
way.

Also if we need to move to Maven for profiles that should be a new task.


was (Author: pferrel):
How does Spark do their multiple version support? They have releases on 1.x and 
2.x because they are not source compatible and I think we are in the same 
situation at least for code that calls the new libs (here I only have 
experience with ES) 

They even maintain versions of the docs site for 1.x and 2.x

We might as well ask these questions now even if we don't want to do it that 
way.



> Cross build for different versions of scala and spark
> -
>
> Key: PIO-30
> URL: https://issues.apache.org/jira/browse/PIO-30
> Project: PredictionIO
>  Issue Type: Improvement
>Reporter: Marcin Ziemiński
>
> The present version of Scala is 2.10 and Spark is 1.4, which is quite old. 
> With Spark 2.0.0 come many performance improvements and features, that people 
> will definitely like to add to their templates. I am also aware that past 
> cannot be ignored and simply dumping 1.x might not be an option for other 
> users. 
> I propose setting up a crossbuild in sbt to build with scala 2.10 and Spark 
> 1.6 and a separate one for Scala 2.11 and Spark 2.0. Most of the files will 
> be consistent between versions including API. The problematic ones will be 
> divided between additional source directories: src/main/scala-2.10/ and 
> src/main/scala-2.11/. The dockerized tests should also take the two versions 
> into consideration



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

[jira] [Comment Edited] (PIO-30) Cross build for different versions of scala and spark

[Pat Ferrel (JIRA)]

[ 
https://issues.apache.org/jira/browse/PIO-30?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15449714#comment-15449714
 ] 

Pat Ferrel edited comment on PIO-30 at 8/30/16 6:16 PM:


How does Spark do their multiple version support? They have releases on 1.x and 
2.x because they are not source compatible and I think we are in the same 
situation at least for code that calls the new libs (here I only have 
experience with ES) 

They even maintain versions of the docs site for 1.x and 2.x

We might as well ask these questions now even if we don't want to do it that 
way.




was (Author: pferrel):
How does Spark do their multiple version support? They even maintain versions 
of the docs site for past versions. We might as well ask these questions now.

> Cross build for different versions of scala and spark
> -
>
> Key: PIO-30
> URL: https://issues.apache.org/jira/browse/PIO-30
> Project: PredictionIO
>  Issue Type: Improvement
>Reporter: Marcin Ziemiński
>
> The present version of Scala is 2.10 and Spark is 1.4, which is quite old. 
> With Spark 2.0.0 come many performance improvements and features, that people 
> will definitely like to add to their templates. I am also aware that past 
> cannot be ignored and simply dumping 1.x might not be an option for other 
> users. 
> I propose setting up a crossbuild in sbt to build with scala 2.10 and Spark 
> 1.6 and a separate one for Scala 2.11 and Spark 2.0. Most of the files will 
> be consistent between versions including API. The problematic ones will be 
> divided between additional source directories: src/main/scala-2.10/ and 
> src/main/scala-2.11/. The dockerized tests should also take the two versions 
> into consideration



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

[jira] [Commented] (PIO-30) Cross build for different versions of scala and spark

[Pat Ferrel (JIRA)]

[ 
https://issues.apache.org/jira/browse/PIO-30?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15449705#comment-15449705
 ] 

Pat Ferrel commented on PIO-30:
---

Fine with me if we can do with cross-build but it's not just different 
libraries it will be different PIO code. There will need to be 2 binaries, and 
2 sets of PIO source.

> Cross build for different versions of scala and spark
> -
>
> Key: PIO-30
> URL: https://issues.apache.org/jira/browse/PIO-30
> Project: PredictionIO
>  Issue Type: Improvement
>Reporter: Marcin Ziemiński
>
> The present version of Scala is 2.10 and Spark is 1.4, which is quite old. 
> With Spark 2.0.0 come many performance improvements and features, that people 
> will definitely like to add to their templates. I am also aware that past 
> cannot be ignored and simply dumping 1.x might not be an option for other 
> users. 
> I propose setting up a crossbuild in sbt to build with scala 2.10 and Spark 
> 1.6 and a separate one for Scala 2.11 and Spark 2.0. Most of the files will 
> be consistent between versions including API. The problematic ones will be 
> divided between additional source directories: src/main/scala-2.10/ and 
> src/main/scala-2.11/. The dockerized tests should also take the two versions 
> into consideration



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

[jira] [Commented] (PIO-32) Create component upgrade release branch

[Pat Ferrel (JIRA)]

[ 
https://issues.apache.org/jira/browse/PIO-32?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15449615#comment-15449615
 ] 

Pat Ferrel commented on PIO-32:
---

Might as well add new ES 2.x compatibility since we will break backwards binary 
compatibility with other upgrades

> Create component upgrade release branch
> ---
>
> Key: PIO-32
> URL: https://issues.apache.org/jira/browse/PIO-32
> Project: PredictionIO
>  Issue Type: New Feature
>Affects Versions: future release
>Reporter: Pat Ferrel
> Fix For: future release
>
>
> Create a release branch for component upgrades that break binary 
> compatibility like Spark 2.x, Scala 2.11, Elasticsearch 2.x



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

[jira] [Created] (PIO-33) Add support for Elasticsearch 2.x to upgrade release branch

[Pat Ferrel (JIRA)]

Pat Ferrel created PIO-33:
-

 Summary: Add support for Elasticsearch 2.x to upgrade release 
branch
 Key: PIO-33
 URL: https://issues.apache.org/jira/browse/PIO-33
 Project: PredictionIO
  Issue Type: New Feature
Affects Versions: future release
Reporter: Pat Ferrel
Assignee: Pat Ferrel
 Fix For: future release


We have code that does this but the remaining issure is that ES encodes data 
differently than 1.x and so teh current form of engine.json `sparkConf` does 
not work. This can be removed as a feature afaik since all of thos params can 
be passed on the command line. Also we could escape illegal characters in the 
engine.json and unescape before passing them to Spark.

This will be binary incompatible with ES 1.x so adding to the upgrade branch



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

[jira] [Commented] (PIO-32) Create component upgrade release branch

[Pat Ferrel (JIRA)]

[ 
https://issues.apache.org/jira/browse/PIO-32?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15449604#comment-15449604
 ] 

Pat Ferrel commented on PIO-32:
---

may be other things to add to this upgrade release branch

> Create component upgrade release branch
> ---
>
> Key: PIO-32
> URL: https://issues.apache.org/jira/browse/PIO-32
> Project: PredictionIO
>  Issue Type: New Feature
>Affects Versions: future release
>Reporter: Pat Ferrel
> Fix For: future release
>
>
> Create a release branch for component upgrades that break binary 
> compatibility like Spark 2.x, Scala 2.11, Elasticsearch 2.x



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

Apache PIO v0.10.0 release

[Pat Ferrel]

It’s been over a week waiting for the template donation paperwork, which I 
imagine will take a few weeks to go through repo creation, license validation 
and PR merging before they can be released. Since they can be on a separate 
release schedules I’d like to remove them from being blocking release. We can 
point the Gallery at Chan’s github for now (just to remove the issue as a 
blocker). Any other opinions?

#3 below is still the main blocker IMO. I have some time this week to handle 
the source license check but would have no clue about how to make SBT include 
the correct licenses in the binary artifacts. Donald is the only SBT expert I 
know in the project can anyone else help with the binary build bits mentioned 
in https://issues.apache.org/jira/browse/PIO-26?

There are several PRs that are in limbo since we aren’t adding new features or 
major new code. The only way to get moving on these is to release. Sooo...

What do people think about a release? What sort of target date can we set? Can 
we get things ready by first week of September?

-- 

What do people think remains for release?

1) template donation and mods. Chan Lee has done work on this but we can’t 
review until the donation and repos are set up.
2) install.sh. There are some suggestions on how to deal with the one-line 
install here: https://issues.apache.org/jira/browse/PIO-22 and a bug here 
https://issues.apache.org/jira/browse/PIO-25 PIO-22 suggests we have an install 
based on source pull and build so it will work even on snapshots in the 
“develop” branch but we could also have an install from binary that works after 
the release. Comments welcome.
3) https://issues.apache.org/jira/browse/PIO-26 has a PR but I don’t fee 
qualified to merge it, can someone pick this up?

Anything else? I took the liberty of marking anything I thought was a 
non-blocker but unresolved as minor. Feel free to disagree. 

Hopefully when #1 comes through we will have everything else cleared up. Let’s 
ship-it.

Re: Regarding PredictionIO templates

[Pat Ferrel]

The current plan is to have a new Gallery which anyone can submit templates to. 
The templates might be hosted on your github account. This needs to get a 
thumbs up from an Apache PIO committer so filing a Jira with your submission 
request should get our attention. If it meets Gallery standards you can create 
a PR (pull request) to the Gallery page with your template description. Then in 
short order your template will be on the Gallery.

To submit for inclusion in Apache is more rigorous requiring that some 
committer is willing to give it long term support and that you assign 
copyrights as per normal Apache procedure.


On Aug 30, 2016, at 2:17 AM, Bansari Shah  wrote:

Thank you, Pat.
I am starting to communicate with Chan to see if more is yet to done. And
we are also working with preditionIO and making our own template for
sentiment analysis. Is there a way to contribute new templates in apache
version of predicitonIO ? We are highly using predictionIO for our
MachineLearninig platform so please update me for any changes or feature
addition in predictionIO.
My github ID : bansarishah

Thank you

Regards,
Bansari Shah


On Tue, Aug 30, 2016 at 2:21 AM, Pat Ferrel  wrote:

> Hi Bansari,
> 
> Chan Lee has done this too, and added tests from the new integration test
> framework. We are awaiting a donation from Saleforce and new repo creation
> before anything can be pushed.
> 
> If you would like to help check work being done on this here:
> https://issues.apache.org/jira/browse/PIO-24
> 
> You might want to coordinate with Chan to see if more is yet to be done,
> for instance checking licenses to make sure all dependencies have
> compatible licenses as described here: https://issues.apache.org/
> jira/browse/PIO-27
> 
> Templates are always released in source form so the binary licenses don’t
> need to be checked.
> 
> 
> On Aug 29, 2016, at 7:17 AM, Bansari Shah  wrote:
> 
> Hey,
> I have updated some of the prediction IO templates from io.predictionio to
> org.apache.predictionio namespace . Is it possible to contribute this
> templates ? If yes can you please guide me how to contribute.
> 
> Thank you
> 
> Regards,
> Bansari Shah
> 
> 

Re: Regarding PredictionIO templates

[Bansari Shah]

Hi Chan,

As you have updated predictionIO templates to org.apache.predictionIO
namespace, is there any thing yet to be done where i can contribute. If
there then please suggest me.

As i am using predictionIO templates for sentiment analysis, i have updated
some templates like coreNLP_SentimentAnalysis template and
OpenNLP_SentimentAnalysis template to org.apache.predictionio namespace. Is
there a way to contribute this templates ?

Please suggest me.

Thank you

Regards,
Bansari Shah

On Tue, Aug 30, 2016 at 2:47 PM, Bansari Shah 
wrote:

> Thank you, Pat.
> I am starting to communicate with Chan to see if more is yet to done. And
> we are also working with preditionIO and making our own template for
> sentiment analysis. Is there a way to contribute new templates in apache
> version of predicitonIO ? We are highly using predictionIO for our
> MachineLearninig platform so please update me for any changes or feature
> addition in predictionIO.
> My github ID : bansarishah
>
> Thank you
>
> Regards,
> Bansari Shah
>
>
> On Tue, Aug 30, 2016 at 2:21 AM, Pat Ferrel  wrote:
>
>> Hi Bansari,
>>
>> Chan Lee has done this too, and added tests from the new integration test
>> framework. We are awaiting a donation from Saleforce and new repo creation
>> before anything can be pushed.
>>
>> If you would like to help check work being done on this here:
>> https://issues.apache.org/jira/browse/PIO-24
>>
>> You might want to coordinate with Chan to see if more is yet to be done,
>> for instance checking licenses to make sure all dependencies have
>> compatible licenses as described here: https://issues.apache.org/jira
>> /browse/PIO-27
>>
>> Templates are always released in source form so the binary licenses don’t
>> need to be checked.
>>
>>
>> On Aug 29, 2016, at 7:17 AM, Bansari Shah 
>> wrote:
>>
>> Hey,
>> I have updated some of the prediction IO templates from io.predictionio to
>> org.apache.predictionio namespace . Is it possible to contribute this
>> templates ? If yes can you please guide me how to contribute.
>>
>> Thank you
>>
>> Regards,
>> Bansari Shah
>>
>>
>