Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Repo info
Activity
    Emmanuel Benazera
    @beniz
    OK, FTR, the answer seems to be here: dmlc/xgboost#834 -> uni-dimentional target.
    Benjamin Pryke
    @benpryke
    I installed the xgboost python package on Win 10 but it only imports successfully if I run python in a terminal with admin privileges. Otherwise, I get "WindowsError: [Error 126] The specified module could not be found" loading libxgboost.dll. Any idea why?
    Boris-Chantel
    @Boris-Chantel
    Hey guys
    Anybody here experienced with hadoop?
    lesshaste
    @lesshaste
    hi... is this room used?
    Dirceu Semighini Filho
    @dirceusemighini
    Hello, is there a plan do add a predict method that receives RDD[LabeledPoint] in XGBoost spark package?
    DestroyNWO
    @DestroyNWO_twitter
    Hey there! anyone used xgboost for ranking. Documentation doesnt provide what is the output of predict in that case. Any idea?
    geoHeil
    @geoHeil
    Hi! How can I get xgboost to predict probabilities instead of a binary outcome for binary classification? http://datascience.stackexchange.com/questions/14527/xgboost-predict-probabilities sklearns predict_proba() does not really seem to work
    Gregory Nwosu
    @gregnwosu
    does this work woth stack?
    Yar Ki
    @yarki
    Has anyone succeed in running xgboost4j? (still struggling with java.lang.UnsatisfiedLinkError)
    geoHeil
    @geoHeil
    Sure works fine. Do the tests run through?
    Yar Ki
    @yarki
    master branch:
    mvn -DskipTests -Dcheckstyle.skip install works fine
    mvn -Dcheckstyle.skip install fails with java.lang.UnsatisfiedLinkError: ml.dmlc.xgboost4j.java.XGBoostJNI.XGDMatrixCreateFromFile(Ljava/lang/String;I[J)I←[0m
    It is a bit unclear where libxgboost4j.dll is supposed to come from
    geoHeil
    @geoHeil
    For me money install works with all tests green.
    Can you compile the container code separately? Is the correct output produced?
    Yar Ki
    @yarki
    Hmm... I'm not sure I got your point
    Yar Ki
    @yarki

    @geoHeil How did you manage to compile native JNI library?

    As far as I see https://github.com/dmlc/xgboost/blob/master/jvm-packages/xgboost4j/src/main/java/ml/dmlc/xgboost4j/java/NativeLibLoader.java expects "xgboost4j.so"/"xgboost4j.dll" to be available.

    Yar Ki
    @yarki
    see https://github.com/dmlc/xgboost/blob/master/jvm-packages/xgboost4j/pom.xml
    it looks like create_jni.bat / create_jni.sh are supposed to copy the image in order to include it into final xgboost4j-0.7.jar
    "For me money install works with all tests green." - are you using a particular tag or a fresh code from the master branch?
    geoHeil
    @geoHeil
    For me checking that https://github.com/dmlc/xgboost/blob/master/doc/build.md runs fine always helped to solve your problem
    No the latest master
    Yar Ki
    @yarki
    Windows / Linux?
    geoHeil
    @geoHeil
    Windows Linux and Mac ;)
    Linux is easier though to get it to work
    Myles Daniel Baker
    @mydpy
    Hi - I am trying to use import ml.dmlc.xgboost4j.scala.spark.XGBoost and getting a strange error when executing XGBoost.train:
    val xgboostModel = XGBoost.train(trainRDD, paramMap, 1, 4, useExternalMemory=true)
    java.lang.NoSuchMethodError: scala.Predef$.ArrowAssoc(Ljava/lang/Object;)Ljava/lang/Object;
        at ml.dmlc.xgboost4j.scala.spark.XGBoost$.overrideParamMapAccordingtoTaskCPUs(XGBoost.scala:227)
        at ml.dmlc.xgboost4j.scala.spark.XGBoost$.trainWithRDD(XGBoost.scala:283)
        at ml.dmlc.xgboost4j.scala.spark.XGBoost$.train(XGBoost.scala:213)
    I compiled for spark version 2.0.1.
    Train is recognized:
    <console>:59: error: not enough arguments for method train: (trainingData: org.apache.spark.rdd.RDD[org.apache.spark.ml.feature.LabeledPoint], params: Map[String,Any], round: Int, nWorkers: Int, obj: ml.dmlc.xgboost4j.scala.ObjectiveTrait, eval: ml.dmlc.xgboost4j.scala.EvalTrait, useExternalMemory: Boolean, missing: Float)ml.dmlc.xgboost4j.scala.spark.XGBoostModel.
    Unspecified value parameters trainingData, params, round, ...
           val xgboostModel = XGBoost.train()
    geoHeil
    @geoHeil
    Did you skip the testcases during compilation of the jvm package?
    Myles Daniel Baker
    @mydpy
    @geoHeil The problem was definitely in compilation and it has been corrected. I'm really not sure what I was doing wrong though, but the examples you referenced ended up working. Managing Spark + XGboost versions is a bit tedius when running on something like EMR, Databricks, etc.
    Thank you for your help!
    geoHeil
    @geoHeil
    np
    skywalkerytx
    @skywalkerytx
    hello guyss, is there a way to set label weight for xgb? for example in a 0-1 problem i really don't care about those = 0 and want to be very sensitive with 1
    Peter Rudenko
    @petro-rudenko
    @skywalkerytx you can use sample_weight parameter, giving more weight to positive classes.
    skywalkerytx
    @skywalkerytx
    @petro-rudenko does this sum(sample_weight_array) need to be ==1?
    Peter Rudenko
    @petro-rudenko
    @skywalkerytx no - the more weight a sample has - the bigger gradient would be.
    skywalkerytx
    @skywalkerytx
    :clap:
    geoHeil
    @geoHeil
    How can I access it sample_weight in python? Just as a regular param like n_estimators for the param map?
    Henry Saputra
    @hsaputra
    HI Peeps
    I am trying to compile XGBoost and use the python-package in my Mac but cannot get rid of the XGBoostLibraryNotFound: Cannot find XGBoost Libarary in the candicate path, did you install compilers and run build.sh in root path? error
    Been following the tips from https://github.com/dmlc/xgboost/blob/master/python-package/build_trouble_shooting.md page but can get it work
    Henry Saputra
    @hsaputra
    seemed like missing OpenMP in my machine, will try to install it and redo ...
    Henry Saputra
    @hsaputra
    Yay, it is working =)
    Holger Peters
    @HolgerPeters
    hi
    I recently reported a bug dmlc/xgboost#1995 and provided a fix with PR dmlc/xgboost#1996
    was wondering if there's anything else to do for a contribution (like writing to a mailing list etc)?
    Denis M Korzhenkov
    @denkorzh
    Hi!
    I've opened an issue #2140 to provide a possibility to have row id in prediction file. Unfortunately it's not a priority now.
    Are there any enthusiasts ready to develope this option for CLI mode?
    Priyanka Goyal
    @goyalpri
    I am working on a big data analytics project, which has frequent updates. I have to perform a lot of analytical queries. Can u guys suggest me the tech stack, I should use for the project?
    I am thinking HBase and Hadoop but since I'm new to big data world, I'm kind of confused. Thanks in advance.
    geoHeil
    @geoHeil
    How big is big? How fast do you need to process updates? Streaming 7 realtime (sub second/ minutes) or batch queries?
    Priyanka Goyal
    @goyalpri
    ~1 inserts/queries per minute.
    and ~2 million db records.