/* Java based auto-weka call sequence */

        // init object and set options
        AutoWEKAClassifier awc = new AutoWEKAClassifier();
        Experiment exp1 = new autoweka.Experiment();

        // set mem limit
        awc.setMemLimit(1024*7);
        memlim = awc.getMemLimit();
        System.out.printf("memlim = %d\n", memlim);
        // set time limit
        awc.setTimeLimit(30);

        // set classifiers
        autoweka.Experiment.AllowedClassifiers("RandomForest");

        // train classifier
        awc.buildClassifier(train);

        // print predictions 
        PrintWriter writer = new PrintWriter(output_file, "UTF-8");
        writer.println("prediction,p1,p2");
        for(i=0; i<test.numInstances(); i++){
            double label = awc.classifyInstance(test.instance(i));
            double[] prob = awc.distributionForInstance(test.instance(i));
            test.instance(i).setClassValue(label);
            writer.printf("%s,", test.instance(i).stringValue(0));
            writer.printf("%f,", prob[0]);
            writer.printf("%f\n", prob[1]);
        }


/* Python based, auto-scikit call sequence */

import autosklearn.classification;
cls = autosklearn.classification.AutoSklearnClassifier(time_left_for_this_task=300, ml_memory_limit=7000, include_estimators=["random_forest", ]);
#include_estimators=["gradient_boosting", ]);
#include_estimators=["adaboost", ]);
cls.fit(X_train, y_train);
y_pred_askl = cls.predict(X_test)