def main() -> None:
input_dir, judgement_year, dividing_point = parse_arguments()
files_to_be_processed = get_files_to_be_processed(input_dir)
numbers, references_number, detriment_words_number = process_files(
files_to_be_processed, input_dir, judgement_year)
exercise1(numbers, judgement_year)
exercise2(numbers, dividing_point, judgement_year)
exercise3(references_number, judgement_year)
exercise4(detriment_words_number, judgement_year)
def main(cl_args):
args = parse_arguments(cl_args)
run_all(
args.emb_models,
args.data_paths,
args.folds,
args.sizes,
args.classifiers,
args.ptv_names,
)
def main():
input_dir, judgement_year = parse_arguments()
create_index_with_analyzer()
files = get_files_to_be_processed(input_dir)
for file in tqdm(files, mininterval=15, unit='files'):
extract_and_upload_data(file, judgement_year, INDEX_DATA_URL, HEADERS)
save_data(
'Word {} occurred {} times in judgements from year {}.'.format(
DETRIMENT_WORD, count_detriment_words(), judgement_year),
'exercise-6.txt')
save_data(
'Given phrase has occurred {} times in judgements from year {}.'.
format(find_phrase(PHRASE, 0), judgement_year), 'exercise-7.txt')
save_data(
'Given phrase has occurred {} times in judgements from year {}.'.
format(find_phrase(PHRASE, 2), judgement_year), 'exercise-8.txt')
save_data(search_top_judges(NUMBER_OF_JUDGES, judgement_year),
'exercise-9.txt')
create_bar_chart(prepare_bar_chart_data(judgement_year), 'exercise-10.png',
judgement_year)
-- Interspersed Sharing Task with Four Hidden Layers
"""
import os
import sys
sys.path.append(os.path.abspath(os.path.dirname(__file__) + '/' + '../..'))
from Experiments.experiment import Experiment
from Models.interspersed_sharing_four_hidden import InterspersedSharingModel
from utils.argument_parser import parse_arguments
from utils.data_utils.data_handler import fetch_data
from utils.training_utils.task_set import Coupled
EXPERIMENT_NAME = 'multi-interspersed-final'
if __name__ == '__main__':
args = parse_arguments()
# Target tasks.
task = Coupled.tasks[args.task_type]
to_run = {('%s-%s' % (EXPERIMENT_NAME, args.task_type)): task}
# These are the training sizes which we will test.
training_sizes = [
500, 1000, 2000, 3000, 4000, 5000, 7500, 10000, 15000, 25000
]
for name, tasks in to_run.iteritems():
# Produce the training, validation, and test sets.
x_train, x_validate, x_test, y_train, y_validate, y_test, task_ids = fetch_data(
tasks)
from MlMerchant import MLMerchant
from abstract_merchant import AbstractMerchant
from ml_engines.rand_for import RandomForestEngine
from utils.argument_parser import parse_arguments
from utils.cross_validator import CrossValidator
from utils.settingsbuilder import SettingsBuilder
class RandomForestMerchant(AbstractMerchant):
def get_cross_validator(self, settings):
return CrossValidator(settings, RandomForestEngine())
def start_merchant(self):
settings = SettingsBuilder() \
.with_data_file('rand_for_models.pkl') \
.build()
ml_merchant = MLMerchant(settings, RandomForestEngine())
ml_merchant.initialize()
return ml_merchant
if __name__ == "__main__":
args = parse_arguments('PriceWars Merchant doing Random Forest Regression')
if args.train and args.buy and args.merchant and args.test and args.output:
RandomForestMerchant().start_cross_validation(args)
else:
RandomForestMerchant().start_server(args)
from MlMerchant import MLMerchant
from abstract_merchant import AbstractMerchant
from ml_engines.log_reg import LogisticRegressionEngine
from utils.argument_parser import parse_arguments
from utils.cross_validator import CrossValidator
from utils.settingsbuilder import SettingsBuilder
class LogisticRegressionMerchant(AbstractMerchant):
def get_cross_validator(self, settings):
return CrossValidator(settings, LogisticRegressionEngine())
def start_merchant(self):
settings = SettingsBuilder() \
.with_data_file('log_reg_models.pkl') \
.build()
ml_merchant = MLMerchant(settings, LogisticRegressionEngine())
ml_merchant.initialize()
return ml_merchant
if __name__ == "__main__":
args = parse_arguments('PriceWars Merchant doing Logistic Regression')
if args.train and args.buy and args.merchant and args.test and args.output:
LogisticRegressionMerchant().start_cross_validation(args)
else:
LogisticRegressionMerchant().start_server(args)
# Sort predictions according to IDs
test_ids, y_pred = zip(*sorted(zip(ids_prediction, predictions)))
else:
# Visualize the histogram of data without outliers
if params['visualize']:
plot_distribution(input_data, 'input_data', verbose=True)
# Initialize weight vector coefficients as zeros
initial_w = np.zeros(input_data.shape[1])
# Train the model
w, loss = least_squares_sgd(yb,
input_data,
initial_w,
batch_size=params['batch_size'],
max_iters=params['max_iters'],
gamma=params['gamma'],
loss_function=params['loss_function'],
verbose=params['verbose'])
y_pred = predict_labels(w, test_data)
# Create a CSV with the predictions either if it was by splitting jets or not
create_csv_submission(test_ids, y_pred, results_path + '/results.csv')
if __name__ == '__main__':
main(**vars(parse_arguments()))
from MlMerchant import MLMerchant
from abstract_merchant import AbstractMerchant
from ml_engines.mlp import MlpEngine
from utils.argument_parser import parse_arguments
from utils.cross_validator import CrossValidator
from utils.settingsbuilder import SettingsBuilder
class MlpMerchant(AbstractMerchant):
def get_cross_validator(self, settings):
return CrossValidator(settings, MlpEngine())
def start_merchant(self):
settings = SettingsBuilder() \
.with_data_file('mlp_models.pkl') \
.build()
ml_merchant = MLMerchant(settings, MlpEngine())
ml_merchant.initialize()
return ml_merchant
if __name__ == "__main__":
args = parse_arguments('PriceWars Merchant doing MLP Regression')
if args.train and args.buy and args.merchant and args.test and args.output:
MlpMerchant().start_cross_validation(args)
else:
MlpMerchant().start_server(args)