def step_save(self, operation: Save):
"""
Perform save operation selected in plan.
:param Save operation: save definition
"""
save_class = save_dict[operation.algorithm]
project_tuple = ProjectTuple(
file_path="",
image=self.image,
segmentation=self.segmentation,
additional_layers=self.additional_layers,
mask=self.mask,
history=self.history,
algorithm_parameters=self.algorithm_parameters,
)
save_path = get_save_path(operation, self.calculation)
save_class.save(save_path, project_tuple, operation.values)
def get_project_info(self) -> ProjectTuple:
algorithm_name = self.last_executed_algorithm
if algorithm_name:
algorithm_val = {
"algorithm_name": algorithm_name,
"values": deepcopy(self.get(f"algorithms.{algorithm_name}")),
}
else:
algorithm_val = {}
return ProjectTuple(
file_path=self.image.file_path,
image=self.image.substitute(),
roi=self.roi,
roi_info=self.roi_info,
additional_layers=self.additional_layers,
mask=self.mask,
history=self.history[:self.history_index + 1],
algorithm_parameters=algorithm_val,
)
def step_save(self, operation: Save):
"""
Perform save operation selected in plan.
:param Save operation: save definition
"""
save_class = save_dict[operation.algorithm]
project_tuple = ProjectTuple(
file_path="",
image=self.image,
roi_info=self.roi_info,
additional_layers=self.additional_layers,
mask=self.mask,
history=self.history,
algorithm_parameters=self.algorithm_parameters,
)
save_path = get_save_path(operation, self.calculation)
save_dir = os.path.dirname(save_path)
os.makedirs(save_dir, exist_ok=True)
save_class.save(save_path, project_tuple, operation.values)
def get_project_info(self) -> ProjectTuple:
algorithm_name = self.last_executed_algorithm
if algorithm_name:
value = self.get(f"algorithms.{algorithm_name}")
if isinstance(value, EventedDict):
value = value.as_dict_deep()
algorithm_val = {
"algorithm_name": algorithm_name,
"values": deepcopy(value),
}
else:
algorithm_val = {}
return ProjectTuple(
file_path=self.image.file_path,
image=self.image.substitute(),
roi_info=self.roi_info,
additional_layers=self.additional_layers,
mask=self.mask,
history=self.history[:self.history_index + 1],
algorithm_parameters=algorithm_val,
points=self.points,
)
def load(
cls,
load_locations: typing.List[typing.Union[str, BytesIO, Path]],
range_changed: typing.Callable[[int, int], typing.Any] = None,
step_changed: typing.Callable[[int], typing.Any] = None,
metadata: typing.Optional[dict] = None,
):
"""Load project from archive old format"""
file_ob: typing.Union[str, tarfile.TarFile, TextIOBase, BufferedIOBase,
RawIOBase, IOBase] = load_locations[0]
if isinstance(file_ob, tarfile.TarFile):
tar_file = file_ob
file_path = ""
elif isinstance(file_ob, str):
tar_file = tarfile.open(file_ob)
file_path = file_ob
elif isinstance(file_ob,
(TextIOBase, BufferedIOBase, RawIOBase, IOBase)):
tar_file = tarfile.open(fileobj=file_ob)
file_path = ""
else:
raise ValueError(f"wrong type of file_ argument: {type(file_ob)}")
image_buffer = BytesIO()
image_tar = tar_file.extractfile(tar_file.getmember("image.npy"))
image_buffer.write(image_tar.read())
image_buffer.seek(0)
image_arr = np.load(image_buffer)
try:
res_buffer = BytesIO()
res_tar = tar_file.extractfile(tar_file.getmember("res_mask.npy"))
res_buffer.write(res_tar.read())
res_buffer.seek(0)
seg_array = np.load(res_buffer)
except KeyError:
seg_array = None
algorithm_str = tar_file.extractfile("data.json").read()
algorithm_dict = json.loads(algorithm_str)
spacing = np.array(
algorithm_dict["spacing"][::-1]) / UNIT_SCALE[Units.nm.value]
image = Image(image_arr.reshape((1, ) + image_arr.shape + (1, )),
spacing, file_path)
values = {
"channel": 0,
"minimum_size": algorithm_dict["minimum_size"],
"threshold": {
"name": "Manual",
"values": {
"threshold": algorithm_dict["threshold"]
}
},
"noise_removal": {
"name": "Gauss",
"values": {
"gauss_type": DimensionType.Layer,
"radius": 1.0
}
} if algorithm_dict["use_gauss"] else {
"name": "None",
"values": {}
},
"side_connection": True,
}
algorithm_parameters = {
"name":
"Upper threshold" if algorithm_dict["threshold_type"] == "Upper"
else "Lower threshold",
"values":
values,
"threshold_list":
algorithm_dict["threshold_list"],
}
return ProjectTuple(file_path,
image,
seg_array,
algorithm_parameters=algorithm_parameters)