def learning(input):
worker = multiprocessing.cpu_count()
if configserver.get('number_worker') > 0:
worker = configserver.get('number_worker')
queue_classifier = multiprocessing.Queue()
for c in classifier.get_all_classifiers():
queue_classifier.put_nowait(c)
processes = []
failed = []
for i in range(worker):
processes += [multiprocessing.Process(target=_learning_worker, args=(input, queue_classifier))]
for process in processes:
process.start()
for process in processes:
process.join()
if not process.exitcode == 0:
failed += [process]
if len(failed) > 0:
logging.error('{} processes have failed - result might not be complete'.format(len(failed)))
def load_classifyhub_settings():
settings = QSettings()
old_config = configserver.get_config()
configserver.parse_args()
for key in old_config.keys():
if configserver.get(key) == old_config[key]:
# This key wasn't changed by cmd arguments - so try to load it from config
data = configserver.get(key)
if isinstance(data, bool):
configserver.set(
key,
bool(
int(
settings.value('classifyhub/{}'.format(key),
old_config[key]))))
elif isinstance(data, int):
configserver.set(
key,
int(
settings.value('classifyhub/{}'.format(key),
old_config[key])))
elif isinstance(data, str):
configserver.set(
key,
str(
settings.value('classifyhub/{}'.format(key),
old_config[key])))
else:
# Save cmd options so no confusion will occur for user
settings.setValue('classifyhub/{}'.format(key),
configserver.get(key))
def batch(input):
if len(input) == 0:
return []
worker = multiprocessing.cpu_count()
if configserver.get('number_worker') > 0:
worker = configserver.get('number_worker')
queue_input = multiprocessing.Queue()
queue_output = multiprocessing.Queue()
for data in input:
queue_input.put_nowait(data)
processes = []
failed = []
for i in range(worker):
processes += [multiprocessing.Process(target=_batch_worker, args=(queue_input, queue_output))]
for process in processes:
process.start()
result = []
alive = True
# We have to pull all elements out of the queue or else the process might not terminate
# See https://docs.python.org/3/library/multiprocessing.html#programming-guidelines
while alive:
try:
while True:
result += [queue_output.get(True, 1)]
except queue.Empty:
pass
alive = False
for process in processes:
if process.is_alive():
alive = True
break
for process in processes:
process.join()
if not process.exitcode == 0:
failed += [process]
if len(failed) > 0:
logging.error('{} processes have failed - result might not be complete'.format(len(failed)))
# Try again just in case we missed some elements
try:
while True:
result += [queue_output.get(True, 1)]
except queue.Empty:
pass
if len(result) != len(input):
logging.error('Expected {} results, got {} - some results are missing'.format(len(input), len(result)))
return result
def check_stale_lock(lock_path, time=20):
worker = multiprocessing.cpu_count()
if configserver.get('number_worker') > 0:
worker = configserver.get('number_worker')
lock = lockfile.LockFile(lock_path)
test_lock = lockfile.LockFile(lock_path + '_STALE')
while not test_lock.i_am_locking():
try:
try:
# Because this is longer then the lock timeout, we are sure that this is stale
test_lock.acquire(timeout=time + time / 5 * worker +
(time / 2) * random.random())
except lockfile.LockTimeout:
test_lock.break_lock()
except Exception as e:
logging.debug('Error at locking stale test log: {}'.format(e))
pass
try:
lock.acquire(timeout=time)
lock.release()
except lockfile.LockTimeout:
logging.debug('Breaking log {}'.format(lock_path))
try:
lock.break_lock()
except lockfile.NotLocked:
pass
try:
test_lock.release()
except lockfile.NotLocked:
# Someone broke it - not bad
pass
def main():
data = processor.file_to_input(configserver.get('input'))
if len(data) == 0:
logging.error('No learning data - aborting')
return
logging.log(configserver.output_log_level(), 'Batch processing started')
result = processor.batch(data)
processor.result_to_file(result, configserver.get('output'))
logging.log(configserver.output_log_level(), 'Batch processing finished')
def check_user_and_secret():
github_secret = github.GithubSecret()
if not github_secret.secret_available:
QMessageBox.warning(
None, 'No authentification possible',
'Authentificated requests are not possible - this will limit you to only a few GitHub requests per hour. '
'Please ensure that the user file ({}) and the secret file ({}) exist in your setup. '
'The file names must match exactly, including the file extension, which is not visible on all systems.'
.format(configserver.get('user_file'),
configserver.get('secret_file')))
def _learning_handle(self):
learning_data = processor.dir_to_learning(
configserver.get('learning_input'))
if len(learning_data) == 0:
QMessageBox.critical(
None, 'No learning data',
'Learning data not found ({}), the application will not work. Please rerun the learning once the configuration was fixed.'
.format(configserver.get('learning_input')))
processor.learning(learning_data)
self._running = False
self.runningChanged.emit(False)
self._learning_running = False
self.learningRunningChanged.emit(False)
def __init__(self, name):
##
# \var config
# \brief Holds the configuration.
#
# As default the conf variable holds an empty dict which can be used, however it is possible to replace
# it with anything that can be serialised as an JSON (e.g. list).
self.config = dict()
if name is '':
logging.warning('Empty name')
name = 'UNKNOWN'
self._dir = configserver.get('model_path')
self._path = self._dir + '/' + name + '.model'
if os.path.isdir(self._dir) and os.path.exists(self._path):
utility.check_stale_lock(self._path + '_LOCK')
lock = lockfile.LockFile(self._path + '_LOCK')
with lock:
with open(self._path, 'r') as file:
try:
self.config = json.load(file)
except:
logging.error('Can not load model {}'.format(self._path))
def main():
data = processor.dir_to_learning(configserver.get('learning_input'))
if len(data) == 0:
logging.error('No learning data - aborting')
return
logging.log(configserver.output_log_level(), 'Learning started')
logging.log(
configserver.output_log_level(),
'Depending on your system, the size of learning data and the amount that needs to be downloaded this might take a while. Please wait.'
)
processor.learning(data)
logging.log(configserver.output_log_level(), 'Learning finished')
def getBoolConfig(self, key):
return bool(
int(
self.settings.value('classifyhub/{}'.format(key),
configserver.get(key))))
def getStringConfig(self, key):
return str(
self.settings.value('classifyhub/{}'.format(key),
configserver.get(key)))
QApplication.setOrganizationName('Top-Ranger')
load_classifyhub_settings()
qmlRegisterType(UIProxy, 'UIProxy', 1, 0, 'UIProxy')
qmlRegisterType(SettingsProxy, 'SettingsProxy', 1, 0, 'SettingsProxy')
# Catch any python error to give an error message
try:
# test for secret
check_user_and_secret()
# Test if we need to learn
need_to_learn = True
if os.path.exists(configserver.get('model_path')):
for file in os.listdir(configserver.get('model_path')):
if os.path.isfile(configserver.get('model_path') + '/' +
file) and file.endswith('.model'):
need_to_learn = False
break
if need_to_learn:
QMessageBox.information(
None, 'No models',
'It seems that no models are present on your system ({}), this means that the learning process has not been run yet.\n'
'The learning process will be started now. This might take some time.'
.format(configserver.get('model_path')))
_learning_needed = True
view = QQmlApplicationEngine()
def main():
# Open Output file
file = None
try:
file = open(configserver.get('output'), 'w')
except OSError:
logging.error('Can not save results to {}'.format(
configserver.get('output')))
# Prepare data
data = processor.dir_to_learning(configserver.get('learning_input'))
if len(data) == 0:
logging.error('No learning data - aborting')
return
k_fold = configserver.get('k-fold')
if k_fold < 2:
logging.error('k-cross must be at least 2 (is: {})'.format(k_fold))
return
logging.log(configserver.output_log_level(),
'Starting validation ({}-cross validation)'.format(k_fold))
logging.log(
configserver.output_log_level(),
'Depending on your system, the size of learning/validation data and the amount that needs to be downloaded this might take a while. Please wait.'
)
if file is not None:
file.write(
'Starting validation ({}-cross validation)\n'.format(k_fold))
file.flush()
datasets = [[] for i in range(k_fold)]
for d in data:
datasets[random.randint(0, k_fold - 1)] += [d]
# Run k-fold cross-validation
precision = utility.get_zero_class_dict()
recall = utility.get_zero_class_dict()
for run in range(k_fold):
logging.log(configserver.output_log_level(),
'Starting validation run {}'.format(run + 1))
if file is not None:
file.write('Starting validation run {}\n'.format(run + 1))
file.flush()
learn = []
truth = []
# Create datasets for run
for i in range(k_fold):
if i == run:
truth = datasets[i]
else:
learn += datasets[i]
# Remove labels
validate = [x[0] for x in truth]
# Learn
processor.learning(learn)
# Calculate validation data set
result = processor.batch(validate)
# Cache results of this run
for c in utility.get_classes():
precision_result = calculate_precision(truth, result, c)
recall_result = calculate_recall(truth, result, c)
if file is not None:
file.write(
'{:6} - precision: {:6.4f}, recall: {:6.4f}\n'.format(
c, precision_result, recall_result))
precision[c] += precision_result
recall[c] += recall_result
if file is not None:
file.write('\n')
file.flush()
# Calculate average
for c in utility.get_classes():
precision[c] /= k_fold
recall[c] /= k_fold
# Print results
logging.log(
configserver.output_log_level(),
'Average results from {}-fold cross-validation:'.format(k_fold))
precision_avg = 0.0
recall_avg = 0.0
if file is not None:
file.write(
'Average results from {}-fold cross-validation:\n'.format(k_fold))
for c in utility.get_classes():
precision_avg += precision[c]
recall_avg += recall[c]
logging.log(
configserver.output_log_level(),
'{:6} - precision: {:6.4f}, recall: {:6.4f}'.format(
c, precision[c], recall[c]))
if file is not None:
file.write('{:6} - precision: {:6.4f}, recall: {:6.4f}\n'.format(
c, precision[c], recall[c]))
precision_avg /= len(utility.get_classes())
recall_avg /= len(utility.get_classes())
logging.log(
configserver.output_log_level(),
'{:6} - precision: {:6.4f}, recall: {:6.4f}'.format(
'ALL', precision_avg, recall_avg))
# Close file if open
if file is not None:
file.write('{:6} - precision: {:6.4f}, recall: {:6.4f}\n'.format(
'ALL', precision_avg, recall_avg))
file.write('\n')
file.close()
