find_best_attributes.py 5.8 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180
  1. # main imports
  2. import os
  3. import sys
  4. import argparse
  5. import pandas as pd
  6. import numpy as np
  7. import logging
  8. import datetime
  9. # model imports
  10. from sklearn.model_selection import train_test_split
  11. from sklearn.model_selection import GridSearchCV
  12. from sklearn.linear_model import LogisticRegression
  13. from sklearn.ensemble import RandomForestClassifier, VotingClassifier
  14. import sklearn.svm as svm
  15. from sklearn.utils import shuffle
  16. from sklearn.externals import joblib
  17. from sklearn.metrics import roc_auc_score
  18. from sklearn.model_selection import cross_val_score
  19. # modules and config imports
  20. sys.path.insert(0, '') # trick to enable import of main folder module
  21. import custom_config as cfg
  22. import models as mdl
  23. from optimization.algorithms.IteratedLocalSearch import IteratedLocalSearch as ILS
  24. from optimization.solutions.BinarySolution import BinarySolution
  25. from optimization.operators.mutators.SimpleMutation import SimpleMutation
  26. from optimization.operators.mutators.SimpleBinaryMutation import SimpleBinaryMutation
  27. from optimization.operators.crossovers.SimpleCrossover import SimpleCrossover
  28. from optimization.operators.policies.RandomPolicy import RandomPolicy
  29. # variables and parameters
  30. models_list = cfg.models_names_list
  31. number_of_values = 26
  32. ils_iteration = 10000
  33. ls_iteration = 20
  34. # default validator
  35. def validator(solution):
  36. if list(solution.data).count(1) < 5:
  37. return False
  38. return True
  39. # init solution (26 attributes)
  40. def init():
  41. return BinarySolution([], 26).random(validator)
  42. def loadDataset(filename):
  43. ########################
  44. # 1. Get and prepare data
  45. ########################
  46. dataset_train = pd.read_csv(filename + '.train', header=None, sep=";")
  47. dataset_test = pd.read_csv(filename + '.test', header=None, sep=";")
  48. # default first shuffle of data
  49. dataset_train = shuffle(dataset_train)
  50. dataset_test = shuffle(dataset_test)
  51. # get dataset with equal number of classes occurences
  52. noisy_df_train = dataset_train[dataset_train.iloc[:, 0] == 1]
  53. not_noisy_df_train = dataset_train[dataset_train.iloc[:, 0] == 0]
  54. nb_noisy_train = len(noisy_df_train.index)
  55. noisy_df_test = dataset_test[dataset_test.iloc[:, 0] == 1]
  56. not_noisy_df_test = dataset_test[dataset_test.iloc[:, 0] == 0]
  57. nb_noisy_test = len(noisy_df_test.index)
  58. final_df_train = pd.concat([not_noisy_df_train[0:nb_noisy_train], noisy_df_train])
  59. final_df_test = pd.concat([not_noisy_df_test[0:nb_noisy_test], noisy_df_test])
  60. # shuffle data another time
  61. final_df_train = shuffle(final_df_train)
  62. final_df_test = shuffle(final_df_test)
  63. # use of the whole data set for training
  64. x_dataset_train = final_df_train.iloc[:,1:]
  65. x_dataset_test = final_df_test.iloc[:,1:]
  66. y_dataset_train = final_df_train.iloc[:,0]
  67. y_dataset_test = final_df_test.iloc[:,0]
  68. return x_dataset_train, y_dataset_train, x_dataset_test, y_dataset_test
  69. def main():
  70. parser = argparse.ArgumentParser(description="Train and find best filters to use for model")
  71. parser.add_argument('--data', type=str, help='dataset filename prefix (without .train and .test)')
  72. parser.add_argument('--choice', type=str, help='model choice from list of choices', choices=models_list)
  73. args = parser.parse_args()
  74. p_data_file = args.data
  75. p_choice = args.choice
  76. print(p_data_file)
  77. # load data from file
  78. x_train, y_train, x_test, y_test = loadDataset(p_data_file)
  79. # create `logs` folder if necessary
  80. if not os.path.exists(cfg.logs_folder):
  81. os.makedirs(cfg.logs_folder)
  82. logging.basicConfig(format='%(asctime)s %(message)s', filename='logs/%s.log' % p_data_file.split('/')[-1], level=logging.DEBUG)
  83. # define evaluate function here (need of data information)
  84. def evaluate(solution):
  85. start = datetime.datetime.now()
  86. # get indices of filters data to use (filters selection from solution)
  87. indices = []
  88. for index, value in enumerate(solution.data):
  89. if value == 1:
  90. indices.append(index)
  91. # keep only selected filters from solution
  92. x_train_filters = x_train.iloc[:, indices]
  93. y_train_filters = y_train
  94. x_test_filters = x_test.iloc[:, indices]
  95. model = mdl.get_trained_model(p_choice, x_train_filters, y_train_filters)
  96. y_test_model = model.predict(x_test_filters)
  97. test_roc_auc = roc_auc_score(y_test, y_test_model)
  98. end = datetime.datetime.now()
  99. diff = end - start
  100. print("Evaluation took :", divmod(diff.days * 86400 + diff.seconds, 60))
  101. return test_roc_auc
  102. # prepare optimization algorithm
  103. updators = [SimpleBinaryMutation(), SimpleMutation(), SimpleCrossover()]
  104. policy = RandomPolicy(updators)
  105. print("Start running ILS")
  106. algo = ILS(init, evaluate, updators, policy, validator, True)
  107. bestSol = algo.run(ils_iteration, ls_iteration)
  108. # print best solution found
  109. print("Found ", bestSol)
  110. # save model information into .csv file
  111. if not os.path.exists(cfg.results_information_folder):
  112. os.makedirs(cfg.results_information_folder)
  113. filename_path = os.path.join(cfg.results_information_folder, cfg.optimization_attributes_result_filename)
  114. filters_counter = 0
  115. # count number of filters
  116. for index, item in enumerate(bestSol.data):
  117. if index != 0 and index % 2 == 1:
  118. # if two attributes are used
  119. if item == 1 or bestSol.data[index - 1] == 1:
  120. filters_counter += 1
  121. line_info = p_data_file + ';' + str(ils_iteration) + ';' + str(ls_iteration) + ';' + str(bestSol.data) + ';' + str(list(bestSol.data).count(1)) + ';' + str(filters_counter) + ';' + str(bestSol.fitness())
  122. with open(filename_path, 'a') as f:
  123. f.write(line_info + '\n')
  124. print('Result saved into %s' % filename_path)
  125. if __name__ == "__main__":
  126. main()