def generate_time_file(include, exclude, output_name='results/data_norm.time'):
files_list = common.get_file_names(include, exclude)
time_dict = {}
data_file = open(output_name, 'w')
data_file.write(
'size\tdimension\tdelta\tembedding time\tl1 time\tedit_time\ttotal time\n'
)
''' format = size, dimension, delta, embedding time, l1 time, edit_time, total time'''
for files in files_list:
file_name = 'distances/{}'.format(files)
dist = numpy.load(file_name)
params = file_name[:-9].split('_')
if len(params) >= 5:
[size, dimension, delta] = map(float, params[2:5])
if dimension <= 20 and size > 2**dimension:
size = 2**dimension
if 'arr_1' in dist.keys():
temp = dist['arr_1'].tolist()
time_dict[files] = temp
norm_factor = size * (size - 1) / 2
str_to_write = '{}\t{}\t{}\t{}\t{}\t{}\t{}\n'.format(
size, dimension, delta, temp['embedding_time'] / size,
temp['l1_distance_time'] / norm_factor,
temp['edit_distance_time'] / norm_factor,
temp['total_time'] / norm_factor)
data_file.write(str_to_write)
data_file.close()
def comp_distortion_variation(include, exclude, delta='0.1'):
files = common.get_file_names(include, exclude)
for file_name in files:
means = []
ans = mean_distortion('distances/' + file_name)
means.append(ans[0])
return means
def generate_time_file(include, exclude, output_name='results/data_norm.time'):
files_list = common.get_file_names(include, exclude)
time_dict = {}
data_file = open(output_name, 'w')
data_file.write('size\tdimension\tdelta\tembedding time\tl1 time\tedit_time\ttotal time\n')
''' format = size, dimension, delta, embedding time, l1 time, edit_time, total time'''
for files in files_list:
file_name = 'distances/{}'.format(files)
dist = numpy.load(file_name)
params = file_name[:-9].split('_')
if len(params) >= 5:
[size, dimension, delta] = map(float, params[2:5])
if dimension <= 20 and size > 2**dimension:
size = 2**dimension
if 'arr_1' in dist.keys():
temp = dist['arr_1'].tolist()
time_dict[files] = temp
norm_factor = size * (size - 1) / 2
str_to_write = '{}\t{}\t{}\t{}\t{}\t{}\t{}\n'.format(size, dimension, delta, temp['embedding_time'] / size, temp[
'l1_distance_time'] / norm_factor, temp['edit_distance_time'] / norm_factor, temp['total_time'] / norm_factor)
data_file.write(str_to_write)
data_file.close()
