def test_find_files():
path = os.path.join(os.getcwd(),'tests','tests_file_io')
expect = [os.path.join(os.getcwd(),'tests','tests_file_io', 'test_find_files.py')]
assert(find_files(substring='test_find_files', path=path) == expect)
# find recursive
assert(find_files(substring='test_find_files.py', path=os.getcwd(), recursive=True) == expect)
# find files and check extension
assert(find_files(substring='test_find_files.py', path=path, check_ext='.py') == expect)
assert(find_files(substring='test_find_files.py', path=path, check_ext='.txt') == [])
def get_index_in_folder(input_filepath, input_basename, folder):
"""Loop through all .dat files in a folder, calling
`get_index_from_filename` on each file"""
path = os.path.join(input_filepath, folder)
files = find_files(input_basename, path=path, check_ext='.dat')
idxs = []
for f in files:
idxs.append(get_index_from_filename(f))
return idxs
def test_find_files():
path = os.path.join(os.getcwd(), 'tests', 'tests_file_io')
expect = [
os.path.join(os.getcwd(), 'tests', 'tests_file_io',
'test_find_files.py')
]
assert (find_files(substring='test_find_files', path=path) == expect)
# find recursive
assert (find_files(substring='test_find_files.py',
path=os.getcwd(),
recursive=True) == expect)
# find files and check extension
assert (find_files(substring='test_find_files.py',
path=path,
check_ext='.py') == expect)
assert (find_files(substring='test_find_files.py',
path=path,
check_ext='.txt') == [])
def get_results_from_training(path):
files = find_files('training', path=path, recursive=True, check_ext='.log')
if len(files) == 0:
return None
header = [
'network', 'n_nodes', 'extra_dense', 'upper_cutoff', 'lower_cutoff',
'smoothing', 'basename', 'best_val_loss', 'best_val_acc', 'best_epoch',
'holdout_acc', 'holdout_loss'
]
missing_holdout = None
result_dict = {el: list() for el in header}
for logpath in tqdm(files):
#do analysis
try:
df_log = pd.read_csv(logpath)
except IOError:
tqdm.write("Can't read: " + logpath)
continue
idx = find_best_epoch(df_log)
best_val_loss, best_val_acc = df_log.loc[idx, ['val_loss', 'val_acc']]
log_filename = os.path.basename(logpath)
basename, _ = os.path.splitext(log_filename)
params = basename.split('__')[:-1]
basename = '__'.join(params)
params.append(basename)
params.append(best_val_loss)
params.append(best_val_acc)
params.append(idx)
try:
# Get holdout performance
holdout_acc, holdout_loss = get_holdout_performance(logpath)
params.append(holdout_acc)
params.append(holdout_loss)
except:
if missing_holdout is None:
tqdm.write("At least one trace missing holdout performance")
missing_holdout = True
continue
for head, par in zip(header, params):
result_dict[head].append(par)
return result_dict
def test_ignore_substring():
path = os.path.join(os.getcwd(), 'tests', 'tests_file_io')
assert (find_files(substring='test_find_files',
path=path,
ignore_substring='test_find') == [])
out = subprocess.call(['sphinx-apidoc', '-o', rst_path, '/Users/sebastian/github/mlxtend/mlxtend', '-e', '-f', '--full', '--follow-links', '--maxdepth', '8'])
# Convert reStructuredText to markdown
files = [(os.path.join(rst_path, f), os.path.join(mkd_path, os.path.splitext(f)[0] + '.md')) for f in os.listdir(rst_path) if f.endswith('.rst')]
for f in files:
out = subprocess.call(['pandoc', f[0], '--from=rst', '--to=markdown', '-o', f[1]])
# Get submodules
files = [f for f in os.listdir(mkd_path) if f.endswith('.md') and len(f.split('.'))>2]
for f in files:
print("- ['api/%s', '%s']" % (f, os.path.splitext(f)[0]))
'''
py_files = find_files(substring='',
path=mlxtend_path,
check_ext='.py',
recursive=True)
py_files = [p for p in py_files if not '__init__.py' in p]
print(py_files)
for f in py_files:
with open(f, 'r') as fd:
file_contents = fd.read()
module = ast.parse(file_contents)
function_definitions = [
node for node in module.body if isinstance(node, ast.FunctionDef)
]
class_definitions = [
node for node in module.body if isinstance(node, ast.ClassDef)
]
def test_ignore_substring():
path = os.path.join(os.getcwd(),'tests','tests_file_io')
assert(find_files(substring='test_find_files', path=path, ignore_substring='test_find') == [])
os.remove
out = subprocess.call(['sphinx-apidoc', '-o', rst_path, '/Users/sebastian/github/mlxtend/mlxtend', '-e', '-f', '--full', '--follow-links', '--maxdepth', '8'])
# Convert reStructuredText to markdown
files = [(os.path.join(rst_path, f), os.path.join(mkd_path, os.path.splitext(f)[0] + '.md')) for f in os.listdir(rst_path) if f.endswith('.rst')]
for f in files:
out = subprocess.call(['pandoc', f[0], '--from=rst', '--to=markdown', '-o', f[1]])
# Get submodules
files = [f for f in os.listdir(mkd_path) if f.endswith('.md') and len(f.split('.'))>2]
for f in files:
print("- ['api/%s', '%s']" % (f, os.path.splitext(f)[0]))
'''
py_files = find_files(substring='', path=mlxtend_path, check_ext='.py', recursive=True)
py_files = [p for p in py_files if not '__init__.py' in p]
print(py_files)
for f in py_files:
with open(f, 'r') as fd:
file_contents = fd.read()
module = ast.parse(file_contents)
function_definitions = [node for node in module.body if isinstance(node, ast.FunctionDef)]
class_definitions = [node for node in module.body if isinstance(node, ast.ClassDef)]
for fu in function_definitions:
print(ast.get_docstring(fu))
for cl in class_definitions:
print(ast.get_docstring(cl))
if not args.input:
print('Please provide an input directory via the -i flag. \n')
args.help()
quit()
if not args.output:
print('Please provide an input directory via the -o flag. \n')
args.help()
quit()
if not os.path.isdir(args.output):
os.mkdir(args.output)
mol2_files = find_files(substring='',
path=args.input,
recursive=args.recursive,
check_ext='.mol2',
ignore_invisible=True)
pbar = pyprind.ProgBar(len(mol2_files))
for path in mol2_files:
mol2_name = os.path.basename(path)
if args.recursive:
case_dir = os.path.split(os.path.split(path)[0])[1]
new_dir = os.path.join(args.output, case_dir)
else:
