Using Fregata to realize spark logical regression algorithm _spark

Import Fregata.spark.data.LibSvmReader Import fregata.spark.metrics.classification. {Areaunderroc, accuracy} import fregata.spark.model.classification.LogisticRegression import Org.apache.spark.
  {sparkconf, sparkcontext}/** * Created by all on 2016/12/8. */Object Fregatafirsttest {def main (args:array[string]): unit = {val conf=new sparkconf (). Setappname ("Test"). Set Master ("local") Val sc=new sparkcontext (conf)//Read data val (_,traindata) =libsvmreader.read (SC,/fre) via the Fregata API GATADATA/A9A ", 123) Val (_,testdata) =libsvmreader.read (SC,"/FREGATADATA/A9A.T ", 123)//Use training data to build a logical regression model Val model= Logisticregression.run (Traindata)///Use test data prediction category Val Pd=model.classpredict (testData)//Evaluate model Val by AUC or other metrics
      Acc=accuracy.of (pd.map{case (x,l), (p,c) => c-> L}) Val Auc=areaunderroc.of (pd.map{ Case ((x,l), (p,c)) => p-> L}) Val loss = Fregata.spark.loss.log (pd.map{case ((x,l), (P,C))
    =>    if (L = = 1d) {(l,c,p)}else{(L, C, 1-p)}}) println (s) areaunderro c = $auc ") println (S" accuracy = $ACC ") println (S" logloss = $loss ")}}

The API for Fregata training data requires a data type of rdd[(Fregata. Vector,fregata. Num)], the predicted API needs to be the same as the training data type or Rdd[fregata. Vector]

The data format