Re: Update default build targets

> For Hadoop 2.6 and Spark 2.1, our updated dependencies will work. +1 We should always catch up latest versions of Hadoop, Spark and so on, but default build targets should cover existing popular environments as much as possible. In addition, HBase version (0.98.5) looks much old. It's already

Re: Update default build targets

Supporting the latest and requiring them are 2 different things. Requiring them (except for ES) means PIO won’t run unless the clusters for every user are upgraded to match the client because only backward compatibility is supported. Last time I checked if you require HDFS 2.7, PIO won’t run on

Re: Update default build targets

These upgrades are very similar to the dependencies we support/provide for PredictionIO 0.11.0-incubating in the Heroku buildpack. If the framework is going to upgrade default dependencies, I wholeheartedly agree that moving to the most recent versions of everything is the way to go. Once PIO

Re: externalizing the spark's es.nodes property from the engine.json

Dan, We use an externalized config with the PredictionIO buildpack for Heroku. If you create a `spark-defaults.conf` file to hold the properties, it will be picked-up by Spark executors, like those passed via `pio train -- --x`. The file should be inside the PredictionIO distribution at:

Re: Error while pio status

Alexey, Can you try to isolate the issue and first check if HBase integration is working? Try to setup PIO to use LOCALFS for model storing: PIO_STORAGE_REPOSITORIES_MODELDATA_SOURCE=LOCALFS And then check if PIO can connect to external HBASE. I have a similar config and my client HBASE

Re: externalizing the spark's es.nodes property from the engine.json

yes, it's UR and the PIO 0.11 Right now it's int engine.json, but I am trying to externalize them so I have the same build in different environments. I don't want to modify and rebuild the template only because the ES host changed. Is there another way except hardcoding it into the engine.json ?

Re: externalizing the spark's es.nodes property from the engine.json

PIO version and template? For the UR, put them in the “sparkConf” section of engine.json named “es.nodes”: “list,of,hosts" comma delimited with no spaces. If not the UR, you may not need this param since the ES master distributes queries in the cluster. On Jun 7, 2017, at 8:42 AM, Dan

Re: Error while pio status

Sorry for the confusion over support, PIO has many components and the docker container you are using is of unknown origin (to me anyway) It seems to have misconfigured something. Please be sure to tell the author or create a PR for it so it can be fixed for other users, it’s one way to pay for

Re: Error while training : NegativeArraySizeException

changing the first / primary / conversion event in eventNames changes what the algorithm will predict. CCO can predict anything in the data by changing this conversion events to the one you want. However that means that you must have good data for the primary/conversion event. Removing it will

Re: Error while pio status

This group is for support of ActionML projects like the Universal Recommender. Please direct PIO questions to the Apache PIO mailing list. On Jun 7, 2017, at 6:32 AM, hed...@gmail.com wrote: I'm trying to setup PIO 0.11.0 in docker using this

Re: Error while training : NegativeArraySizeException

I have a model deployed for this app, it works if I keep only (facet, search) as event types. When asking for a prediction to my deployed model I have an answer in relation with my data (about cars). I checked that the data is sent at the right accesskey in the right ES index. This part is fine I

Re: Error while training : NegativeArraySizeException

Also what version of UR you're into? Is it the latest one? I've only worked with UR 0.50 . On Wed, Jun 7, 2017 at 3:12 PM, Vaghawan Ojha wrote: > Yes you need to build the app again when you change something in the > engine.json. That is every time when you change

Re: Error while training : NegativeArraySizeException

Yes the three event types that I defined in the engine.json exist in my dataset, facet is my primary, I checked that it exists. I think it is not needed to build again when changing something in the engine.json, as the file is read in the process but I built it and tried again and I still have

Re: Error while training : NegativeArraySizeException

Hi, For me this problem had happened when I had mistaken my primary events. The first eventName in the eventName array "eventNames": ["facet","view","search"] is primary. There is that event in your data. Did you make sure, you built the app again when you changed the eventName in engine.json?

Re: Error while training : NegativeArraySizeException

You could explicitly do pio train -- --master spark://localhost:7077 --driver-memory 16G --executor-memory 24G and change the spark master url and the memories configuration. And see if that works. Thanks On Wed, Jun 7, 2017 at 1:55 PM, Bruno LEBON wrote: > Hi, > > Using

Error while training : NegativeArraySizeException

Hi, Using UR with PIO 0.10 I am trying to train my dataset. In return I get the following error: *...* *[INFO] [DataSource] Received events List(facet, view, search)* *[INFO] [DataSource] Number of events List(5, 4, 6)* *[INFO] [Engine$] org.template.TrainingData does not support data sanity