def test_three():
config = utils.load_config()
dermis = inputs.SkinData(config['data_dir'], 'dermis')
with tf.Graph().as_default():
global_step = tf.train.get_or_create_global_step()
image_ph = tf.placeholder(dtype=tf.float32, shape=(None, None, 3))
images = tf.expand_dims(image_ph, axis=0)
net = FCN(images, net_params=config['net_params'])
saver = tf.train.Saver()
with tf.Session() as sess:
saver.restore(sess,
tf.train.latest_checkpoint(config['train_dir']))
logger.info('Model at step-%d restored successfully!' %
sess.run(global_step))
utils.create_if_not_exists(config['save_path'])
for i, (image,
label) in enumerate(zip(dermis.images, dermis.labels)):
if i % 5 == 0:
logger.info('Processing image %d...' % i)
prep_image = image_prep_for_test(image)
pred = np.squeeze(
sess.run(net.outputs, feed_dict={image_ph: prep_image}))
path = os.path.join(config['save_path'],
dermis.listing[i].split('/')[-1] + '.jpg')
save_all(image, label, pred, path)
def get_train_val(train: datasets,
test_transform: transforms,
dataset: str,
val_perc: float = 0.1):
"""
Extract val_perc% of the training set as the validation set.
:param train: training dataset
:param test_transform: transformation of the test dataset
:param dataset: dataset name
:param val_perc: percentage of the training set to be extracted
:return: the training set and the validation set
"""
dataset_length = train.data.shape[0]
directory = 'datasets/val_permutations/'
create_if_not_exists(directory)
file_name = dataset + '.pt'
if os.path.exists(directory + file_name):
perm = torch.load(directory + file_name)
else:
perm = torch.randperm(dataset_length)
torch.save(perm, directory + file_name)
train.data = train.data[perm]
train.targets = np.array(train.targets)[perm]
test_dataset = ValidationDataset(
train.data[:int(val_perc * dataset_length)],
train.targets[:int(val_perc * dataset_length)],
transform=test_transform)
train.data = train.data[int(val_perc * dataset_length):]
train.targets = train.targets[int(val_perc * dataset_length):]
return train, test_dataset
def test_five():
config = utils.load_config()
dermis = inputs.SkinData(config['data_dir'], 'dermis')
with tf.Graph().as_default():
global_step = tf.train.get_or_create_global_step()
image_ph = tf.placeholder(dtype=tf.float32, shape=(None, None, 3))
images = tf.expand_dims(image_ph, axis=0)
net = FCN(images, net_params=config['net_params'])
h, w = tf.shape(image_ph)[0], tf.shape(image_ph)[1]
upscore = tf.image.resize_images(net.endpoints['conv4'], size=(h, w))
prob_one = tf.nn.sigmoid(upscore)
prob_zero = 1 - prob_one
probs = tf.concat([prob_zero, prob_one], axis=3)
saver = tf.train.Saver()
with tf.Session() as sess:
saver.restore(sess,
tf.train.latest_checkpoint(config['train_dir']))
logger.info('Model at step-%d restored successfully!' %
sess.run(global_step))
utils.create_if_not_exists(config['save_path'])
result_before = {'TP': 0, 'TN': 0, 'FP': 0, 'FN': 0}
result_after = result_before.copy()
def update_dict(d, to_update):
for key in to_update:
d[key] += to_update[key]
_, x_test, _, y_test = train_test_split(
dermis.images,
dermis.labels,
random_state=config['split_seed'])
for i, (image, label) in enumerate(zip(x_test, y_test)):
prep_image = image_prep_for_test(image)
probs_o = np.squeeze(
sess.run(probs, feed_dict={image_ph: prep_image}))
cnn_result = np.argmax(probs_o, axis=2)
cnn_crf_result = crf_post_process(image, probs_o)
result_bi = count_many(cnn_result, label)
result_ai = count_many(cnn_crf_result, label)
update_dict(result_before, result_bi)
update_dict(result_after, result_ai)
result_bi.update(metric_many_from_counter(result_bi))
result_ai.update(metric_many_from_counter(result_ai))
if i % 5 == 0:
logger.info(
'Image-{}\nresult before\n{}\nresult after\n{}\n'.
format(i, result_bi, result_ai))
result_before.update(metric_many_from_counter(result_before))
result_after.update(metric_many_from_counter(result_after))
logger.info('\nresult before\n{}\nresult after\n{}\n'.format(
result_before, result_after))
def download_forecast_data(target_date, force=False):
"""Download forecast data for the given target date.
If force is False, data is only downloaded if it doesn't already exist.
"""
start_date = end_date = target_date
target_date_formatted = start_date.strftime(DATE_FORMAT)
folder = os.path.join(settings.DATA_FOLDER, 'MyOcean', 'Forecast')
create_if_not_exists(folder)
datasets = [
{ # chlorophile, nitrate, phosphate, oxygen...
'service': 'http://purl.org/myocean/ontology/service/database#MEDSEA_ANALYSIS_FORECAST_BIO_006_006-TDS',
'product': 'myov04-med-ogs-bio-an-fc',
'time': '12:00:00',
},
{ # salinity
'service': 'http://purl.org/myocean/ontology/service/database#MEDSEA_ANALYSIS_FORECAST_PHYS_006_001_a-TDS',
'product': 'myov05-med-ingv-sal-an-fc-dm',
'time': '00:00:00'
},
{ # temperature
'service': 'http://purl.org/myocean/ontology/service/database#MEDSEA_ANALYSIS_FORECAST_PHYS_006_001_a-TDS',
'product': 'myov05-med-ingv-tem-an-fc-dm',
'time': '00:00:00'
},
{ # temperature
'service': 'http://purl.org/myocean/ontology/service/database#MEDSEA_ANALYSIS_FORECAST_PHYS_006_001_a-TDS',
'product': 'myov05-med-ingv-cur-an-fc-dm',
'time': '00:00:00',
'variables': ['vozocrtx', 'vomecrty']
}
]
for dataset in datasets:
filename = '%s-%s.nc' % (dataset['product'], target_date_formatted)
if not exists(filename, folder) or force:
try:
download_myocean_data(
service=dataset['service'],
product=dataset['product'],
variables=dataset.get('variables'),
time_start='%s %s' % (start_date, dataset['time']),
time_end='%s %s' % (end_date, dataset['time']),
folder=folder,
filename=filename)
except MaxTriesExceededException:
logger.error("Download of file {} from MyOcean failed.".format(
filename))
else:
print(
'File %s already exists, skipping download... (use force=True to override).'
% filename)
def write(self, args: Dict[str, Any]) -> None:
"""
writes out the logged value along with its arguments.
:param args: the namespace of the current experiment
"""
for cc in useless_args:
if cc in args:
del args[cc]
columns = list(args.keys())
new_cols = []
for i, acc in enumerate(self.accs):
args['task' + str(i + 1)] = acc
new_cols.append('task' + str(i + 1))
columns = new_cols + columns
create_if_not_exists(base_path() + "results/" + self.setting)
create_if_not_exists(base_path() + "results/" + self.setting + "/" +
self.dataset)
create_if_not_exists(base_path() + "results/" + self.setting + "/" +
self.dataset + "/" + self.model)
write_headers = False
path = base_path() + "results/" + self.setting + "/" + self.dataset\
+ "/" + self.model + "/mean_accs.csv"
if not os.path.exists(path):
write_headers = True
with open(path, 'a') as tmp:
writer = csv.DictWriter(tmp, fieldnames=columns)
if write_headers:
writer.writeheader()
writer.writerow(args)
if self.setting == 'class-il':
create_if_not_exists(base_path() + "results/task-il/" +
self.dataset)
create_if_not_exists(base_path() + "results/task-il/" +
self.dataset + "/" + self.model)
for i, acc in enumerate(self.accs_mask_classes):
args['task' + str(i + 1)] = acc
write_headers = False
path = base_path() + "results/task-il" + "/" + self.dataset + "/"\
+ self.model + "/mean_accs.csv"
if not os.path.exists(path):
write_headers = True
with open(path, 'a') as tmp:
writer = csv.DictWriter(tmp, fieldnames=columns)
if write_headers:
writer.writeheader()
writer.writerow(args)
def download_forecast_data(target_date, force=False):
"""Download forecast data for the given target date.
If force is False, data is only downloaded if it doesn't already exist.
"""
start_date = end_date = target_date
target_date_formatted = start_date.strftime(DATE_FORMAT)
folder = os.path.join(settings.DATA_FOLDER, 'MyOcean', 'Forecast')
create_if_not_exists(folder)
datasets = [
{ # chlorophile, nitrate, phosphate, oxygen...
'service': 'http://purl.org/myocean/ontology/service/database#MEDSEA_ANALYSIS_FORECAST_BIO_006_006-TDS',
'product': 'myov04-med-ogs-bio-an-fc',
'time': '12:00:00',
},
{ # salinity
'service': 'http://purl.org/myocean/ontology/service/database#MEDSEA_ANALYSIS_FORECAST_PHYS_006_001_a-TDS',
'product': 'myov05-med-ingv-sal-an-fc-dm',
'time': '00:00:00'
},
{ # temperature
'service': 'http://purl.org/myocean/ontology/service/database#MEDSEA_ANALYSIS_FORECAST_PHYS_006_001_a-TDS',
'product': 'myov05-med-ingv-tem-an-fc-dm',
'time': '00:00:00'
},
{ # temperature
'service': 'http://purl.org/myocean/ontology/service/database#MEDSEA_ANALYSIS_FORECAST_PHYS_006_001_a-TDS',
'product': 'myov05-med-ingv-cur-an-fc-dm',
'time': '00:00:00',
'variables': ['vozocrtx', 'vomecrty']
}
]
for dataset in datasets:
filename = '%s-%s.nc' % (dataset['product'], target_date_formatted)
if not exists(filename, folder) or force:
try:
download_myocean_data(
service=dataset['service'],
product=dataset['product'],
variables=dataset.get('variables'),
time_start='%s %s' % (start_date, dataset['time']),
time_end='%s %s' % (end_date, dataset['time']),
folder=folder,
filename=filename)
except MaxTriesExceededException:
logger.error("Download of file {} from MyOcean failed.".format(filename))
else:
print('File %s already exists, skipping download... (use force=True to override).' % filename)
def test_four():
config = utils.load_config()
dermis = inputs.SkinData(config['data_dir'], 'dermis')
with tf.Graph().as_default():
global_step = tf.train.get_or_create_global_step()
image_ph = tf.placeholder(dtype=tf.float32, shape=(None, None, 3))
images = tf.expand_dims(image_ph, axis=0)
net = FCN(images, net_params=config['net_params'])
h, w = tf.shape(image_ph)[0], tf.shape(image_ph)[1]
upscore = tf.image.resize_images(net.endpoints['conv4'], size=(h, w))
prob_one = tf.nn.sigmoid(upscore)
prob_zero = 1 - prob_one
probs = tf.concat([prob_zero, prob_one], axis=3)
saver = tf.train.Saver()
with tf.Session() as sess:
saver.restore(sess,
tf.train.latest_checkpoint(config['train_dir']))
logger.info('Model at step-%d restored successfully!' %
sess.run(global_step))
utils.create_if_not_exists(config['save_path'])
total_count = 0
true_count = 0
for i, (image,
label) in enumerate(zip(dermis.images, dermis.labels)):
prep_image = image_prep_for_test(image)
probs_o = np.squeeze(
sess.run(probs, feed_dict={image_ph: prep_image}))
cnn_result = np.argmax(probs_o, axis=2)
accuracy_before, _, _ = metric_accuracy(cnn_result, label)
cnn_crf_result = crf_post_process(image, probs_o)
accuracy_i, true_count_i, total_count_i = metric_accuracy(
cnn_crf_result, label)
true_count += true_count_i
total_count += total_count_i
ss = 'DOWN' if accuracy_before > accuracy_i else 'UP'
path = os.path.join(config['save_path'],
dermis.listing[i].split('/')[-1])
path = '{:<6} ({:.3f}) ({:.3f}) {}.jpg'.format(
path, accuracy_before, accuracy_i, ss)
save_all_two(image, label, cnn_result, cnn_crf_result, path)
if i % 5 == 0:
logger.info(
'Image-%d accuracy before(%.3f) after(%.3f) %s' %
(i, accuracy_before, accuracy_i, ss))
accuracy = true_count * 1.0 / total_count
logger.info('Accuracy after crf: %.3f' % accuracy)
def predict_forecast(target_date, force=False):
"""Predict the presence of medusae using a previously calibrated model.
If force is False, it is only predicted the output file doesn't already exist.
"""
for a in settings.TUNEZ_DATA:
filename = '{}EF-{}.csv'.format(a, target_date.strftime(DATE_FORMAT))
folder = os.path.join(settings.DATA_FOLDER, 'Projections')
create_if_not_exists(folder)
if not exists(filename, folder) or force:
os.chdir(settings.DATA_FOLDER)
with open(os.path.join(BASE_DIR, 'R', 'Tunisia_MedJellyRisk_{}.R'.format(a)), 'r') as inputfile:
call(["R", "--no-save",
"--args", target_date.strftime(DATE_FORMAT)], stdin=inputfile)
else:
print('\nFile %s already exists, skipping prediction... (use force=True to override).' % filename)
def preprocess_forecast_data(target_date, force=False):
"""Preprocess forecast environmental data from MyOcean using R.
If force is False, it is only preprocessed if the output file doesn't already exist.
"""
filename = 'Forecast_Env-%s.csv' % target_date.strftime(DATE_FORMAT)
folder = os.path.join(settings.DATA_FOLDER, 'MyOcean', 'Forecast')
create_if_not_exists(folder)
if not exists(filename, folder) or force:
os.chdir(os.path.join(settings.DATA_FOLDER))
with open(os.path.join(BASE_DIR, 'R', 'ExtractData_MyOcean.R'), 'r') as inputfile:
call(["R", "--no-save",
"--args", target_date.strftime(DATE_FORMAT)], stdin=inputfile)
else:
print('\nFile %s already exists, skipping preprocessing... (use force=True to override).' % filename)
def preprocess_forecast_data(target_date, force=False):
"""Preprocess forecast environmental data from MyOcean using R.
If force is False, it is only preprocessed if the output file doesn't already exist.
"""
filename = 'Forecast_Env-%s.csv' % target_date.strftime(DATE_FORMAT)
folder = os.path.join(settings.DATA_FOLDER, 'MyOcean', 'Forecast')
create_if_not_exists(folder)
if not exists(filename, folder) or force:
os.chdir(os.path.join(settings.DATA_FOLDER))
with open(os.path.join(BASE_DIR, 'R', 'ExtractData_MyOcean.R'),
'r') as inputfile:
call([
"R", "--no-save", "--args",
target_date.strftime(DATE_FORMAT)
],
stdin=inputfile)
else:
print(
'\nFile %s already exists, skipping preprocessing... (use force=True to override).'
% filename)
def predict_forecast(target_date, force=False):
"""Predict the presence of medusae using a previously calibrated model.
If force is False, it is only predicted the output file doesn't already exist.
"""
for a in settings.TUNEZ_DATA:
filename = '{}EF-{}.csv'.format(a, target_date.strftime(DATE_FORMAT))
folder = os.path.join(settings.DATA_FOLDER, 'Projections')
create_if_not_exists(folder)
if not exists(filename, folder) or force:
os.chdir(settings.DATA_FOLDER)
with open(
os.path.join(BASE_DIR, 'R',
'Tunisia_MedJellyRisk_{}.R'.format(a)),
'r') as inputfile:
call([
"R", "--no-save", "--args",
target_date.strftime(DATE_FORMAT)
],
stdin=inputfile)
else:
print(
'\nFile %s already exists, skipping prediction... (use force=True to override).'
% filename)
def update_backup(commands):
if (commands.virtual_drive):
# Если диск V уже занят, смонтирует диск на указаной букве
virtual_drive = commands.virtual_drive
else:
virtual_drive = DEFAULT_VIRTUAL_DRIVE
if (not virtual_drive.endswith(':')):
virtual_drive = virtual_drive + ':'
if (commands.name):
volume = utils.find_path_to_volume_by_backup_name(
commands.name, program_directory)
elif (commands.volume):
volume = commands.volume
else:
return Font.YELLOW + '[!] Введите имя бэкапа или путь к тому бэкапа для обновления'
if (commands.password):
if (len(commands.password) < MIN_PASSWORD_LENGTH):
return Font.YELLOW + '[!] Пароль слишком короткий. Минимум 25 символов'
else:
return Font.YELLOW + '[!] Пароль не найден'
if (utils.volume_is_mount(virtual_drive)):
print(Font.YELLOW +
'[!] Том уже смонтирован или диск с таким именем уже существует')
else:
# Монтирование тома
if (not utils.mount_veracrypt_volume(DEFAULT_VERACRYPT_PATH, volume,
commands.password,
virtual_drive)):
return Font.YELLOW + '[!] Возникла ошибка при монтировании тома'
# Проверка смонтированого тома на наличие нужных файлов бэкапа
if (not utils.is_backup_drive(virtual_drive)):
return Font.YELLOW + '[i] Диск не является бэкапом'
print(Font.YELLOW + '[i] Загрузка старых метаданных...')
try:
backup_metadata = utils.load_metadata_from_json(
os.path.join(virtual_drive, 'metadata.json'))
except utils.CastomException as exc:
return Font.YELLOW + exc
except Exception as exc:
return Font.YELLOW + '[!] Возникла непредвиденая ошибка: %s' % exc
metadata = backup_metadata['metadata']
backup_name = backup_metadata['backup_name']
backup_directory = backup_metadata['directory']
compression_level = backup_metadata['compression_level']
amount_files_in_backup = backup_metadata['amount_files_in_backup']
last_filelist = utils.read_file(os.path.join(
virtual_drive, 'filelist.txt')) # Загрузка старого списка файлов
if (commands.blacklist):
blacklist = utils.read_file(commands.blacklist)
else:
blacklist = utils.read_file(
os.path.join(virtual_drive,
'blacklist.txt')) # Загрузка старого черного списка
if (commands.recursion_level):
max_recursion_level = commands.recursion_level
else:
max_recursion_level = backup_metadata['recursion_level']
''' ЗАГРУЗКА СТАРЫХ И ПОЛУЧЕНИЕ НОВЫХ ДАННЫХ, ИХ СРАВНИВАНИЕ, ОТСЛЕЖИВАНИЕ ТИПА ИЗМЕНЕНИЙ И РАСПРЕДИЛЕНИЕ ИХ ПО СООТВЕТСТВУЮЩИХ СПИСКАХ'''
print(Font.YELLOW + '[i] Начало сбора данных...')
# Получение списка файлов находящихся в папке бэкапа
new_filelist = collect_backup_files(backup_directory, blacklist,
max_recursion_level)
print(Font.YELLOW + '[i] Найдено файлов: %i шт.' % len(new_filelist))
# Сравнивание списков файло для поика изменений
deleted_files, appended_files = utils.cmp_lists(last_filelist,
new_filelist)
# КОСТЫЛЬ. Удаляем со списка удаленных нашу "подставную папку" посколькуон находиться в директории программы
# а не в директории для архивации, что в свою очередь скрипт расценит это как удаление файла
if (IGNORED_EMPTY_DIR in deleted_files):
deleted_files.remove(IGNORED_EMPTY_DIR)
# Отслеживание типа изменений и также изменение списка удаленных файлов
changes_list = utils.identify_changes(metadata, appended_files,
deleted_files)
# Добавление метаданных новых файлов
for change in changes_list:
parent, filename, status = change
if (not parent and not status):
metadata.update({filename: get_information(filename)})
# Проверка на наличие обновлений только существующих файлов с архива
updated_files = []
for filename in metadata:
if (os.path.exists(filename)):
if (os.stat(filename).st_mtime > metadata[filename]['st_mtime']):
updated_files.append(filename)
metadata.update({filename: get_information(filename)})
if (commands.verbosity):
print(Font.GREEN + '[>] Файл %s нужно обновить' % filename)
for data in changes_list:
if (all(data)):
updated_files.append(data[0])
if (commands.verbosity):
print(Font.GREEN + '[>] Файл %s %s в %s' %
(data[0], CHANGES_STATUSES.get(data[2], 'Неизвестно'),
data[1]))
for file in deleted_files:
if (commands.verbosity):
print(Font.GREEN + '[>] Файл %s был удален!' % file)
'''КОНЕЦ ЭТОГО БЛОКА'''
# Поик копий, перемещений и переименований файлов для создания зависимостей для уменьшения
appended_files = optimize_metadata(changes_list, metadata)
amount_appended_files = len(appended_files)
amount_updated_files = len(updated_files)
amount_deleted_files = len(deleted_files)
amount_changes = amount_updated_files + amount_appended_files + amount_deleted_files
if (amount_changes >= MIN_AMOUNT_CHANGES_TO_UPDATE or commands.force):
if (amount_appended_files > 0):
print(Font.CYAN +
'[i] Добавлено файлов: %i шт.' % amount_appended_files)
if (amount_updated_files > 0):
print(Font.CYAN +
'[i] Обновлено файлов: %i шт.' % amount_updated_files)
if (amount_deleted_files):
print(Font.CYAN +
'[i] Удалено файлов: %i шт.' % amount_deleted_files)
# Создание папки для бэкапа
utils.create_if_not_exists(os.path.join(virtual_drive, 'updates'))
# Если будем делать обновление бэкапа, сохраняем старые данные и список изменений для возможности отката изменений
updates_directory = os.path.join(
virtual_drive, 'updates',
time.ctime(time.time()).replace(':', '-'))
# Проверка наличия папки для сохранения старых метаданных
utils.create_if_not_exists(updates_directory)
# Сохранение, старых метаданных. Зачем? Хз :D
try:
utils.dump_metadata_to_json(
os.path.join(updates_directory, 'metadata.json'),
backup_metadata)
utils.dump_metadata_to_txt(
os.path.join(updates_directory, 'filelist.txt'), last_filelist)
utils.dump_metadata_to_json(
os.path.join(updates_directory, 'changes.json'), changes_list)
except utils.CastomException as exc:
print(Font.YELLOW + exc)
# Обновление файлов в архиве
if (updated_files):
asigned_updated_files = utils.asign_unf(updated_files, metadata)
utils.update_files(virtual_drive, backup_name, compression_level,
asigned_updated_files, Font)
# Добавление новых файлов в архив
if (appended_files):
asigned_appended_files = utils.asign_unf(appended_files, metadata)
utils.compress_files(virtual_drive, backup_name, compression_level,
asigned_appended_files, Font)
# Запись в метаданные метки, что файл удален
if (deleted_files):
utils.set_flags_is_deleted_files(metadata, deleted_files)
backup_metadata.update({
'last_update':
time.ctime(time.time()),
'amount_appended_filse':
amount_appended_files,
'amount_updated_files':
amount_updated_files,
'amount_deleted_files':
amount_deleted_files,
'amount_files_in_backup':
amount_files_in_backup + amount_appended_files -
amount_deleted_files,
'metadata':
metadata,
})
update_backup_metadata(virtual_drive, backup_metadata, new_filelist,
blacklist)
auto_dismount_veracrypt_volume_or_open_backup_drive(
commands, virtual_drive)
return Font.CYAN + '[>] Бэкап успешно обновлен!'
else:
return Font.YELLOW + '[!] Бэкап не требует обовления'
# и получить некоторые файлы, лучше сначала запустить сессию указав имя бэкапа или путь к тому и пароль
# session -p 1234567890123456788901234567890ABCDEF -v "volume.hc"
# иначе при каждом действии нужно будет вводить эти данные
session = {
'--password': None,
'--volume': None,
'--name': None,
'--virtual_drive': None,
}
# Путь к корневой папке програмы, нужен для размещения файла с данными о полным пути к файлу бэкапа и его имени
# Нужно для того, чтобы при обновлении бэкапа не указывать путь к нему, а всего лить имя ПРИМЕР -update -n BACKUP_1
program_directory = os.path.dirname(sys.argv[0]).replace('/', '\\')
#
IGNORED_EMPTY_DIR = os.path.join(program_directory, EMPTY_DIR)
utils.create_if_not_exists(IGNORED_EMPTY_DIR)
#
Font = colorama.Fore if STYLE == 0 else colorama.Back
# Если установлен флажок --o/-open открывает в explorer бэкап и запрещает автоматическое размонтирование
def auto_dismount_veracrypt_volume_or_open_backup_drive(
commands, virtual_drive):
if (commands.open):
print(Font.YELLOW + '[i] Открытие папки бэкапа...')
utils.open_backup_drive(virtual_drive)
else:
print(Font.YELLOW + '[i] Начало размонтирования тома...')
utils.dismount_veracrypt_volume(DEFAULT_VERACRYPT_PATH, virtual_drive)