def save(self, data: Union[torch.Tensor, np.ndarray], meta_data: Optional[Dict] = None) -> None:
"""
Save data into a Nifti file.
The meta_data could optionally have the following keys:
- ``'filename_or_obj'`` -- for output file name creation, corresponding to filename or object.
- ``'original_affine'`` -- for data orientation handling, defaulting to an identity matrix.
- ``'affine'`` -- for data output affine, defaulting to an identity matrix.
- ``'spatial_shape'`` -- for data output shape.
- ``'patch_index'`` -- if the data is a patch of big image, append the patch index to filename.
When meta_data is specified, the saver will try to resample batch data from the space
defined by "affine" to the space defined by "original_affine".
If meta_data is None, use the default index (starting from 0) as the filename.
Args:
data: target data content that to be saved as a NIfTI format file.
Assuming the data shape starts with a channel dimension and followed by spatial dimensions.
meta_data: the meta data information corresponding to the data.
See Also
:py:meth:`monai.data.nifti_writer.write_nifti`
"""
filename = meta_data[Key.FILENAME_OR_OBJ] if meta_data else str(self._data_index)
self._data_index += 1
original_affine = meta_data.get("original_affine", None) if meta_data else None
affine = meta_data.get("affine", None) if meta_data else None
spatial_shape = meta_data.get("spatial_shape", None) if meta_data else None
patch_index = meta_data.get(Key.PATCH_INDEX, None) if meta_data else None
if isinstance(data, torch.Tensor):
data = data.detach().cpu().numpy()
path = create_file_basename(self.output_postfix, filename, self.output_dir, self.data_root_dir, patch_index)
path = f"{path}{self.output_ext}"
# change data shape to be (channel, h, w, d)
while len(data.shape) < 4:
data = np.expand_dims(data, -1)
# change data to "channel last" format and write to nifti format file
data = np.moveaxis(np.asarray(data), 0, -1)
# if desired, remove trailing singleton dimensions
if self.squeeze_end_dims:
while data.shape[-1] == 1:
data = np.squeeze(data, -1)
write_nifti(
data,
file_name=path,
affine=affine,
target_affine=original_affine,
resample=self.resample,
output_spatial_shape=spatial_shape,
mode=self.mode,
padding_mode=self.padding_mode,
align_corners=self.align_corners,
dtype=self.dtype,
output_dtype=self.output_dtype,
)
def filename(self, subject: PathLike = "subject", idx=None, **kwargs):
"""
Create a filename based on the input ``subject`` and ``idx``.
The output filename is formed as:
``output_dir/[subject/]subject[_postfix][_idx][_key-value][ext]``
Args:
subject: subject name, used as the primary id of the output filename.
When a `PathLike` object is provided, the base filename will be used as the subject name,
the extension name of `subject` will be ignored, in favor of ``extension``
from this class's constructor.
idx: additional index name of the image.
kwargs: additional keyword arguments to be used to form the output filename.
The key-value pairs will be appended to the output filename as ``f"_{k}-{v}"``.
"""
full_name = create_file_basename(
postfix=self.postfix,
input_file_name=subject,
folder_path=self.output_dir,
data_root_dir=self.data_root_dir,
separate_folder=self.parent,
patch_index=idx,
makedirs=self.makedirs,
)
for k, v in kwargs.items():
full_name += f"_{k}-{v}"
if self.ext is not None:
ext = f"{self.ext}"
full_name += f".{ext}" if ext and not ext.startswith(".") else f"{ext}"
return full_name
def save(self,
data: Union[torch.Tensor, np.ndarray],
meta_data: Optional[Dict] = None) -> None:
"""
Save data into a Nifti file.
The meta_data could optionally have the following keys:
- ``'filename_or_obj'`` -- for output file name creation, corresponding to filename or object.
- ``'original_affine'`` -- for data orientation handling, defaulting to an identity matrix.
- ``'affine'`` -- for data output affine, defaulting to an identity matrix.
- ``'spatial_shape'`` -- for data output shape.
When meta_data is specified, the saver will try to resample batch data from the space
defined by "affine" to the space defined by "original_affine".
If meta_data is None, use the default index (starting from 0) as the filename.
Args:
data: target data content that to be saved as a NIfTI format file.
Assuming the data shape starts with a channel dimension and followed by spatial dimensions.
meta_data: the meta data information corresponding to the data.
See Also
:py:meth:`monai.data.nifti_writer.write_nifti`
"""
filename = meta_data["filename_or_obj"] if meta_data else str(
self._data_index)
for _ in range(self.output_name_uplevel):
filename = os.path.dirname(filename)
self._data_index += 1
original_affine = meta_data.get("original_affine",
None) if meta_data else None
affine = meta_data.get("affine", None) if meta_data else None
spatial_shape = meta_data.get("spatial_shape",
None) if meta_data else None
if torch.is_tensor(data):
data = data.detach().cpu().numpy()
filename = create_file_basename(self.output_postfix, filename,
self.output_dir)
filename = f"{filename}{self.output_ext}"
# change data shape to be (channel, h, w, d)
while len(data.shape) < 4:
data = np.expand_dims(data, -1)
# change data to "channel last" format and write to nifti format file
data = np.moveaxis(data, 0, -1)
write_nifti(
data,
file_name=filename,
affine=affine,
target_affine=original_affine,
resample=self.resample,
output_spatial_shape=spatial_shape,
mode=self.mode,
padding_mode=self.padding_mode,
align_corners=self.align_corners,
dtype=self.dtype,
)
def save(self,
data: Union[torch.Tensor, np.ndarray],
meta_data: Optional[Dict] = None) -> None:
"""
Save data into a png file.
The meta_data could optionally have the following keys:
- ``'filename_or_obj'`` -- for output file name creation, corresponding to filename or object.
- ``'spatial_shape'`` -- for data output shape.
- ``'patch_index'`` -- if the data is a patch of big image, append the patch index to filename.
If meta_data is None, use the default index (starting from 0) as the filename.
Args:
data: target data content that to be saved as a png format file.
Assuming the data shape are spatial dimensions.
Shape of the spatial dimensions (C,H,W).
C should be 1, 3 or 4
meta_data: the meta data information corresponding to the data.
Raises:
ValueError: When ``data`` channels is not one of [1, 3, 4].
See Also
:py:meth:`monai.data.png_writer.write_png`
"""
filename = meta_data[Key.FILENAME_OR_OBJ] if meta_data else str(
self._data_index)
self._data_index += 1
spatial_shape = meta_data.get(
"spatial_shape", None) if meta_data and self.resample else None
patch_index = meta_data.get(Key.PATCH_INDEX,
None) if meta_data else None
if isinstance(data, torch.Tensor):
data = data.detach().cpu().numpy()
path = create_file_basename(self.output_postfix, filename,
self.output_dir, self.data_root_dir,
patch_index)
path = f"{path}{self.output_ext}"
if data.shape[0] == 1:
data = data.squeeze(0)
elif 2 < data.shape[0] < 5:
data = np.moveaxis(np.asarray(data), 0, -1)
else:
raise ValueError(
f"Unsupported number of channels: {data.shape[0]}, available options are [1, 3, 4]"
)
write_png(
np.asarray(data),
file_name=path,
output_spatial_shape=spatial_shape,
mode=self.mode,
scale=self.scale,
)
def test_value(self):
with tempfile.TemporaryDirectory() as tempdir:
output_tmp = os.path.join(tempdir, "output")
result = create_file_basename("", "test.txt", output_tmp, "")
expected = os.path.join(output_tmp, "test", "test")
self.assertEqual(result, expected)
result = create_file_basename("", os.path.join("foo", "test.txt"), output_tmp, "")
expected = os.path.join(output_tmp, "test", "test")
self.assertEqual(result, expected)
result = create_file_basename("", os.path.join("foo", "test.txt"), output_tmp, "foo")
expected = os.path.join(output_tmp, "test", "test")
self.assertEqual(result, expected)
result = create_file_basename("", os.path.join("foo", "bar", "test.txt"), output_tmp, "foo")
expected = os.path.join(output_tmp, "bar", "test", "test")
self.assertEqual(result, expected)
result = create_file_basename(
postfix="",
input_file_name=os.path.join("foo", "bar", "data", "test.txt"),
folder_path=output_tmp,
data_root_dir=os.path.join("foo", "bar"),
)
expected = os.path.join(output_tmp, "data", "test", "test")
self.assertEqual(result, expected)
result = create_file_basename("", os.path.join("foo", "bar", "test.txt"), output_tmp, "bar")
expected = os.path.join(tempdir, "foo", "bar", "test", "test")
self.assertEqual(result, expected)
result = create_file_basename("", os.path.join("rest", "test.txt"), output_tmp, "rest")
expected = os.path.join(tempdir, "output", "test", "test")
self.assertEqual(result, expected)
result = create_file_basename("", "test.txt", output_tmp, "foo")
expected = os.path.join(output_tmp, "test", "test")
self.assertEqual(result, expected)
result = create_file_basename("post", "test.tar.gz", output_tmp, "foo")
expected = os.path.join(output_tmp, "test", "test_post")
self.assertEqual(result, expected)
def save(self, data: Union[torch.Tensor, np.ndarray], meta_data: Optional[dict] = None) -> None:
"""
Save data into a png file.
The meta_data could optionally have the following keys:
- ``'filename_or_obj'`` -- for output file name creation, corresponding to filename or object.
- ``'spatial_shape'`` -- for data output shape.
If meta_data is None, use the default index (starting from 0) as the filename.
Args:
data: target data content that to be saved as a png format file.
Assuming the data shape are spatial dimensions.
Shape of the spatial dimensions (C,H,W).
C should be 1, 3 or 4
meta_data: the meta data information corresponding to the data.
Raises:
ValueError: PNG image should only have 1, 3 or 4 channels.
See Also
:py:meth:`monai.data.png_writer.write_png`
"""
filename = meta_data["filename_or_obj"] if meta_data else str(self._data_index)
self._data_index += 1
spatial_shape = meta_data.get("spatial_shape", None) if meta_data and self.resample else None
if torch.is_tensor(data):
data = data.detach().cpu().numpy()
filename = create_file_basename(self.output_postfix, filename, self.output_dir)
filename = f"{filename}{self.output_ext}"
if data.shape[0] == 1:
data = data.squeeze(0)
elif 2 < data.shape[0] < 5:
data = np.moveaxis(data, 0, -1)
else:
raise ValueError("PNG image should only have 1, 3 or 4 channels.")
write_png(
data, file_name=filename, output_spatial_shape=spatial_shape, mode=self.mode, scale=self.scale,
)
def test_relative_path(self):
output = create_file_basename("", "test.txt", "output", "", makedirs=False)
expected = os.path.join("output", "test", "test")
self.assertEqual(output, expected)
