Skip to content

Commit

Permalink
Replaced functions with io.convert_to_tomostack
Browse files Browse the repository at this point in the history
  • Loading branch information
@AndrewHerzing
AndrewHerzing committed Oct 2, 2023
1 parent 73c6488 commit 716d63f
Showing 1 changed file with 69 additions and 173 deletions.
242 changes: 69 additions & 173 deletions tomotools/io.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -19,27 +19,25 @@
logger.setLevel(logging.INFO)


def numpy_to_tomo_stack(data, tilts=None, manual_tilts=False):
def convert_to_tomo_stack(data, tilts=None, manual_tilts=False):
"""
Create a TomoStack object from a NumPy array.
This will retain both the axes information and the metadata.
Create a TomoStack object from other data structures.
Parameters
----------
data : Numpy array
Array containing tilt series data. First dimension must represent
the tilt axis. The second and third dimensions are the X and Y
image dimentsions, respectively
data : Numpy array, Hyperspy Signal2D
Tilt series data to be converted. If NumPy array, the first
dimension must represent the tilt dimension. If Hyperspy Signal2D, the
the tilt dimension must be the navigation axis.
tilts : Numpy array
tilts : None, Numpy array, or Hyperpsy Signal1D
Array containing the tilt values in degrees for each projection.
If provided, these values will be stored in
stack.metadata.Tomography.tilts
manual_tilts : bool
If True, prompt for input of maximum positive tilt, maximum negative
tilt, and tilt increment
tilt, and tilt increment in order to population stack.metadata.Tomography.tilts
Returns
Expand All @@ -50,184 +48,82 @@ def numpy_to_tomo_stack(data, tilts=None, manual_tilts=False):
Examples
--------
>>> import numpy as np
>>> from tomotools.io import convert_to_tomo_stack
>>> s = np.random.random((50, 500,500))
>>> from tomotools.io import numpy_to_tomo_stack
>>> s_new = numpy_to_tomo_stack(s)
>>> s_new = convert_to_tomo_stack(s)
>>> s_new
<TomoStack, title: , dimensions: (50|500, 500)>
"""
s = signal_to_tomo_stack(hspy.signals.Signal2D(data), tilts)

s.axes_manager[0].name = 'Tilt'
s.axes_manager[0].units = 'unknown'
s.axes_manager[1].name = 'x'
s.axes_manager[1].units = 'unknown'
s.axes_manager[2].name = 'y'
s.axes_manager[2].units = 'unknown'

if type(tilts) is np.ndarray:
s.metadata.Tomography.tilts = tilts
s.axes_manager[0].units = 'degrees'
s.axes_manager[0].offset = tilts[0]
s.axes_manager[0].scale = tilts[1] - tilts[0]
elif manual_tilts:
negtilt = eval(input('Enter maximum negative tilt: '))
postilt = eval(input('Enter maximum positive tilt: '))
tiltstep = eval(input('Enter tilt step: '))
tilts = np.arange(negtilt, postilt + tiltstep, tiltstep)
s.metadata.Tomography.tilts = tilts
s.axes_manager[0].scale = tilts[1] - tilts[0]
s.axes_manager[0].offset = tilts[0]
s.axes_manager[0].units = 'degrees'

axes_list = [x for _,
x in sorted(s.axes_manager.as_dictionary().items())]
metadata_dict = s.metadata.as_dictionary()
original_metadata_dict = s.original_metadata.as_dictionary()
s = TomoStack(s, axes=axes_list, metadata=metadata_dict,
original_metadata=original_metadata_dict)
return s


def signal_to_tomo_stack(s, tilt_signal=None, manual_tilts=False):
"""
Create a TomoStack object from a HyperSpy signal.
This will retain both the axes information and the metadata.
Parameters
----------
s : HyperSpy Signal2D or BaseSignal
HyperSpy signal to be converted to TomoStack object
tilt_signal : HyperSpy Signal1D or BaseSignal or NumPy array
Signal or array that defines the tilt axis for the signal in degrees.
manual_tilts : bool
If True, prompt for input of maximum positive tilt, maximum negative
tilt, and tilt increment
Returns
-------
tomo_stack_signal : TomoStack object
Examples
--------
>>> import numpy as np
>>> from tomotools.io import convert_to_tomo_stack
>>> import hyperspy.api as hs
>>> s = hs.signals.Signal2D(np.random.random((50, 500,500)))
>>> s.metadata.General.title = "test dataset"
>>> s
<Signal2D, title: test dataset, dimensions: (50|500, 500)>
>>> from tomotools.io import signal_to_tomo_stack
>>> s_new = signal_to_tomo_stack(s)
>>> s_new = convert_to_tomo_stack(s)
>>> s_new
<TomoStack, title: test dataset, dimensions: (50|500, 500)>
"""
s_new = s.deepcopy()
if not s_new.metadata.has_item("Tomography"):
s_new.metadata.add_node("Tomography")

if tilt_signal is not None:
if type(tilt_signal) in [np.ndarray, list]:
s_new.metadata.Tomography.tilts = tilt_signal
s_new.axes_manager[0].name = 'Tilt'
s_new.axes_manager[0].units = 'degrees'
s_new.axes_manager[0].offset = tilt_signal[0]
s_new.axes_manager[0].scale = tilt_signal[1] - tilt_signal[0]
else:
s_new.axes_manager[0].name = tilt_signal.axes_manager[0].name
s_new.axes_manager[0].units = tilt_signal.axes_manager[0].units
s_new.axes_manager[0].scale = tilt_signal.axes_manager[0].scale
s_new.axes_manager[0].offset = tilt_signal.axes_manager[0].offset
<TomoStack, title: , dimensions: (50|500, 500)>
elif manual_tilts:
"""
def _set_axes(s):
s.axes_manager[0].name = 'Tilt'
s.axes_manager[0].units = 'degrees'
s.axes_manager[1].name = 'x'
s.axes_manager[2].name = 'y'
return s

def _set_tomo_metadata(s):
tomo_metadata = {"cropped": False,
"shifts": np.zeros([s.data.shape[0], 2]),
"tiltaxis": 0,
"tilts": np.zeros(s.data.shape[0]),
"xshift": 0,
"yshift": 0}
s.metadata.add_node("Tomography")
s.metadata.Tomography.add_dictionary(tomo_metadata)
s.metadata.Tomography
return s

def _get_manual_tilts():
negtilt = eval(input('Enter maximum negative tilt: '))
postilt = eval(input('Enter maximum positive tilt: '))
tiltstep = eval(input('Enter tilt step: '))
tilts = np.arange(negtilt, postilt + tiltstep, tiltstep)
logger.info('User provided tilts stored')
s_new.axes_manager[0].name = 'Tilt'
s_new.axes_manager[0].units = 'degrees'
s_new.axes_manager[0].scale = tilts[1] - tilts[0]
s_new.axes_manager[0].offset = tilts[0]

elif s.metadata.has_item("Tomography"):
if type(s.metadata.Tomography.tilts) is np.ndarray:
tilts = s_new.metadata.Tomography.tilts
s_new.axes_manager[0].name = 'Tilt'
s_new.axes_manager[0].units = 'degrees'
s_new.axes_manager[0].scale = tilts[1] - tilts[0]
s_new.axes_manager[0].offset = tilts[0]
logger.info("Tilts found in TomoStack metadata")

elif s.metadata.has_item('Acquisition_instrument.TEM.Stage.tilt_alpha'):
tilt_alpha = s.metadata.Acquisition_instrument.TEM.Stage.tilt_alpha
if type(tilt_alpha) is np.ndarray:
n = s.data.shape[0]
tilts = s.metadata.Acquisition_instrument.TEM.Stage.tilt_alpha[0:n]
logger.info('Tilts found in metadata')
s_new.axes_manager[0].name = 'Tilt'
s_new.axes_manager[0].units = 'degrees'
s_new.axes_manager[0].scale = tilts[1] - tilts[0]
s_new.axes_manager[0].offset = tilts[0]

elif s.metadata.has_item('Acquisition_instrument.SEM.Stage.tilt_alpha'):
tilt_alpha = s.metadata.Acquisition_instrument.SEM.Stage.tilt_alpha
if type(tilt_alpha) is np.ndarray:
n = s.data.shape[0]
tilts = s.metadata.Acquisition_instrument.SEM.Stage.tilt_alpha[0:n]
logger.info('Tilts found in metadata')
s_new.axes_manager[0].name = 'Tilt'
s_new.axes_manager[0].units = 'degrees'
s_new.axes_manager[0].scale = tilts[1] - tilts[0]
s_new.axes_manager[0].offset = tilts[0]

elif s.metadata.General.has_item('original_filename'):
tiltfile = ('%s.rawtlt' % os.path.splitext(
s.metadata.General.original_filename)[0])
if os.path.isfile(tiltfile):
tilts = np.loadtxt(tiltfile)
logger.info('.rawtlt file detected.')
s_new.axes_manager[0].name = 'Tilt'
s_new.axes_manager[0].units = 'degrees'
s_new.axes_manager[0].scale = tilts[1] - tilts[0]
s_new.axes_manager[0].offset = tilts[0]
if len(tilts) == s.data.shape[0]:
logger.info('Tilts loaded from .rawtlt file')
else:
logger.info('Number of tilts in .rawtlt file inconsistent'
' with data shape')
return tilts

if type(data) is np.ndarray:
stack = hspy.signals.Signal2D(data)
elif type(data) is hspy.signals.Signal2D:
stack = data.deepcopy()
else:
raise (TypeError, "Unsupported data type. Must be either"
"NumPy Array or Hyperspy Signal")

elif s_new.axes_manager[0].name in ['Tilt', 'Tilts', 'Angle', 'Angles',
'Theta', 'tilt', 'tilts', 'angle',
'angles', 'theta']:
logger.info("Tilts found in HyperSpy signal axis 0")
stack = _set_tomo_metadata(stack)

if type(tilts) is np.ndarray:
pass
elif type(tilts) is hspy.signals.Signal1D:
tilts = tilts.data
elif manual_tilts:
tilts = _get_manual_tilts()
else:
s_new.axes_manager[0].name = 'Tilt'
s_new.axes_manager[0].units = 'unknown'
if s_new.axes_manager[1].name not in ['x', 'X']:
s_new.axes_manager[1].name = 'x'
s_new.axes_manager[1].units = 'unknown'
if s_new.axes_manager[2].name not in ['y', 'Y']:
s_new.axes_manager[2].name = 'y'
s_new.axes_manager[2].units = 'unknown'
logger.info('Tilts not found. Calibrate axis 0')
tilts = None

axes_list =\
[x for _, x in sorted(s_new.axes_manager.as_dictionary().items())]

metadata = s_new.metadata.as_dictionary()
original_metadata = s_new.original_metadata.as_dictionary()

s_new = TomoStack(s.data, axes=axes_list, metadata=metadata,
original_metadata=original_metadata)
return s_new
tilts = np.zeros(stack.data.shape[0])
logger.info('Tilts are note defined. Please add tilts to Tomography metadata.')

if tilts.shape != stack.data.shape[0]:
raise (ValueError, "Number of tilts is not consistent with data shape."
"%s does not equal %s" % (tilts.shape, stack.data.shape))

stack.metadata.Tomography.tilts = tilts
stack.axes_manager[0].offset = tilts[0]
stack.axes_manager[0].scale = tilts[1] - tilts[0]

axes_list = [x for _,
x in sorted(stack.axes_manager.as_dictionary().items())]
metadata_dict = stack.metadata.as_dictionary()
original_metadata_dict = stack.original_metadata.as_dictionary()
stack = TomoStack(stack, axes=axes_list, metadata=metadata_dict,
original_metadata=original_metadata_dict)
return stack


def loadhspy(filename, tilts=None):
Expand Down Expand Up @@ -437,7 +333,7 @@ def load_dm_series(input_data):
stack.axes_manager[0].units = 'degrees'
stack.axes_manager[0].offset = tilts_sorted[0]

stack = signal_to_tomo_stack(stack)
stack = convert_to_tomo_stack(stack)
stack.metadata.Tomography.tilts = tilts_sorted
return stack

Expand Down

0 comments on commit 716d63f

Please sign in to comment.