main_tree.py

import numpy as np
import alg_radmm_simple
import alg_radmm_imgproc_band
import alg_radmm_nmf_decomp
import alg_radmm_nn_signal_noise
import alg_radmm_tree_signal_noise
import  alg_radmm_nmf_decomp_time
from sklearn.model_selection import train_test_split

def main():
    np.random.seed(13)
    #radmmAlg = alg_radmm_simple.AlgRadMMSimple("/mnt/ssd/radiologicalthreatsmm")
    #radmmAlg = alg_radmm_imgproc_band.AlgRadMMImgProcBand("/mnt/ssd/radiologicalthreatsmm")
    #radmmAlg = alg_radmm_nmf_decomp.AlgRadMMNmfDecomp("/mnt/ssd/radiologicalthreatsmm")
    #radmmAlg = alg_radmm_nn_signal_noise.AlgRadMMNeuralNetSignalNoise("/mnt/ssd/radiologicalthreatsmm")
    radmmAlg = alg_radmm_tree_signal_noise.AlgRadMMTreeSignalNoise("/mnt/ssd/radiologicalthreatsmm")
    #radmmAlg = alg_radmm_nmf_decomp_time.AlgRadMMNmfDecompTime("/mnt/ssd/radiologicalthreatsmm")
    
    train_ids = radmmAlg.get_train_ids()
    train_ids_set, val_ids_set = train_test_split(train_ids, random_state=13, test_size=0.20, shuffle=True)
    train_x = radmmAlg.get_train_x(train_ids_set, False)
    train_y = radmmAlg.get_train_y(train_ids_set)
    radmmAlg.train(train_x, train_y, train_ids_set)

#    val_x = radmmAlg.get_train_x(val_ids_set, True)
#    pred_y = radmmAlg.predict(val_x, val_ids_set, export=True, validation=True)
#    print("Score: %f" % radmmAlg.score(pred_y, val_ids_set, verbose=True))
#    radmmAlg.write_submission(pred_y, val_ids_set, "submission.csv")

#    test_ids = radmmAlg.get_test_ids()
#    test_x = radmmAlg.get_test_x(test_ids)
#    pred_y = radmmAlg.predict(test_x, test_ids, validation=False)
#    radmmAlg.write_submission(pred_y, test_ids, "solution.csv")

    #radmmAlg.export_predict_trace(val_ids_set)

if __name__ == '__main__':
    main()