def main():
parser = argparse.ArgumentParser(description='One-hot-code data')
parser.add_argument('input_path', type=str, help='full path to input file')
parser.add_argument('output_path',
type=str,
help='full path to output file')
parser.add_argument('number_max_bins',
type=int,
help='number of max bins for output data')
args = parser.parse_args()
input_path = args.input_path
output_file = args.output_path
num_bins = args.number_max_bins
data_frame = FileManager.get_csv_file_data_pandas(input_path)
print('data frame head:')
print(data_frame.head(3))
hot_coded_data_frame = DataManipulator.one_hot_code(data_frame, num_bins)
print('data one hot coded head:')
print(hot_coded_data_frame.head(3))
hot_coded_data_frame.to_csv(output_file,
header=None,
index=None,
sep=',',
mode='a')
def main():
#print('LOG: Main program to run tests')
parser = argparse.ArgumentParser(
description=
'Run Stepwise Forward Selection && K-Means clustering on given data')
parser.add_argument('file_path', type=str, help='full path to input file')
parser.add_argument('num_clusters',
type=int,
default=2,
nargs='?',
help='number of total clusters')
args = parser.parse_args()
#INPUTS
print()
print('INPUTS')
file_path = args.file_path
print('input file path:', file_path)
num_clusters = args.num_clusters
print('number of clusters: ', num_clusters)
print()
#READ INPUT DATA
input_data = FileManager.get_csv_file_data_pandas(file_path)
print('head of input data')
print(input_data.head())
print()
#PREPROCESS DATA
#According to 'names' files && visual inspection there are 0 missing values for these data sets
#Remove missing value rows
#input_data.pd.dropna(inplace=True)
#print()
#RUN FEATURE SELECTION && K-MEANS
sfs = StepwiseForwardSelection(input_data)
chosen_features, chosen_data_set = sfs.run_sfs(num_clusters)
print()
print('RESULT:')
print('Chosen features (indexed from 0):')
print(chosen_features)
print('Chosen data:')
print(chosen_data_set)
print('Best silhouette coefficient:')
print(sfs.base_performance)
print('Clusters:')
print(sfs.chosen_model.clusters)
print('Centroids:')
print(sfs.chosen_model.centroids)
print()
def split_data(file_path, is_random, num_groups=5, separator=','):
#GET DATA
original_data = FileManager.get_csv_file_data_pandas(file_path, separator)
#print('original data')
#print(original_data)
#STANDARD STUFF
#Split the data into 5 groups
groups = list()
num_groups = int(num_groups)
if (is_random == True):
#Use basic random 5-way split
groups = DataManipulator.split_data_randomly(original_data, num_groups)
else:
#Use more complex split
groups = DataManipulator.split_data_randomly_accounting_for_class(
original_data, num_groups)
return groups
