def transform_scalars(self, dataset):
"""Define this method for Python operators that
transform input scalars"""
from tomviz import utils
import numpy as np
self.progress.maximum = NUMBER_OF_CHUNKS
scalars = utils.get_scalars(dataset)
if scalars is None:
raise RuntimeError("No scalars found!")
if scalars.min() < 0:
print("WARNING: Square root of negative values results in NaN!")
else:
# transform the dataset
# Process dataset in chunks so the user gets an opportunity to
# cancel.
result = np.float32(scalars)
step = 0
for chunk in np.array_split(result, NUMBER_OF_CHUNKS):
if self.canceled:
return
np.sqrt(chunk, chunk)
step += 1
self.progress.value = step
# set the result as the new scalars.
utils.set_scalars(dataset, result)
def transform_scalars(dataset, constant=0):
"""Add a constant to the data set"""
from tomviz import utils
import numpy as np
scalars = utils.get_scalars(dataset)
if scalars is None:
raise RuntimeError("No scalars found!")
# Try to be a little smart so that we don't always just produce a
# double-precision output
newMin = np.min(scalars) + constant
newMax = np.max(scalars) + constant
if (constant).is_integer() and newMin.is_integer() and newMax.is_integer():
# Let ints be ints!
constant = int(constant)
newMin = int(newMin)
newMax = int(newMax)
for dtype in [np.uint8, np.int8, np.uint16, np.int16, np.uint32, np.int32,
np.uint64, np.int64, np.float32, np.float64]:
if np.can_cast(newMin, dtype) and np.can_cast(newMax, dtype):
constant = np.array([constant], dtype=dtype)
break
# numpy should cast to an appropriate output type to avoid overflow
result = scalars + constant
utils.set_scalars(dataset, result)
def transform_scalars(dataset):
"""Reinterpret a signed integral array type as its unsigned counterpart.
This can be used when the bytes of a data array have been interpreted as a
signed array when it should have been interpreted as an unsigned array."""
from tomviz import utils
import numpy as np
scalars = utils.get_scalars(dataset)
if scalars is None:
raise RuntimeError("No scalars found!")
dtype = scalars.dtype
dtype = dtype.type
typeMap = {np.int8: np.uint8, np.int16: np.uint16, np.int32: np.uint32}
typeAddend = {np.int8: 128, np.int16: 32768, np.int32: 2147483648}
if dtype not in typeMap:
raise RuntimeError("Scalars are not int8, int16, or int32")
newType = typeMap[dtype]
addend = typeAddend[dtype]
newScalars = scalars.astype(dtype=newType) + addend
utils.set_scalars(dataset, newScalars)
def transform_scalars(dataset):
"""Reinterpret a signed integral array type as its unsigned counterpart.
This can be used when the bytes of a data array have been interpreted as a
signed array when it should have been interpreted as an unsigned array."""
from tomviz import utils
import numpy as np
scalars = utils.get_scalars(dataset)
if scalars is None:
raise RuntimeError("No scalars found!")
dtype = scalars.dtype
dtype = dtype.type
typeMap = {
np.int8: np.uint8,
np.int16: np.uint16,
np.int32: np.uint32
}
typeAddend = {
np.int8: 128,
np.int16: 32768,
np.int32: 2147483648
}
if dtype not in typeMap:
raise RuntimeError("Scalars are not int8, int16, or int32")
newType = typeMap[dtype]
addend = typeAddend[dtype]
newScalars = scalars.astype(dtype=newType) + addend
utils.set_scalars(dataset, newScalars)
def transform_scalars(self, dataset):
"""Define this method for Python operators that
transform input scalars"""
from tomviz import utils
import numpy as np
self.progress.maximum = NUMBER_OF_CHUNKS
scalars = utils.get_scalars(dataset)
if scalars is None:
raise RuntimeError("No scalars found!")
if scalars.min() < 0:
print("WARNING: Square root of negative values results in NaN!")
else:
# transform the dataset
# Process dataset in chunks so the user gets an opportunity to
# cancel.
result = np.float32(scalars)
step = 0
for chunk in np.array_split(result, NUMBER_OF_CHUNKS):
if self.canceled:
return
np.sqrt(chunk, chunk)
step += 1
self.progress.value = step
# set the result as the new scalars.
utils.set_scalars(dataset, result)
def transform_scalars(dataset):
from tomviz import utils
import numpy as np
scalars = utils.get_scalars(dataset)
if scalars is None:
raise RuntimeError("No scalars found!")
result = np.amax(scalars) - scalars
utils.set_scalars(dataset, result)
def transform_scalars(dataset):
from tomviz import utils
import numpy as np
scalars = utils.get_scalars(dataset)
if scalars is None:
raise RuntimeError("No scalars found!")
result = np.amax(scalars) - np.float32(scalars)
utils.set_scalars(dataset, result)
def transform_scalars(dataset):
"""Define this method for Python operators that
transform input scalars"""
from tomviz import utils
import numpy as np
scalars = utils.get_scalars(dataset)
if scalars is None:
raise RuntimeError("No scalars found!")
# transform the dataset
result = np.sqrt(scalars)
# set the result as the new scalars.
utils.set_scalars(dataset, result)
def transform_scalars(dataset):
"""Define this method for Python operators that
transform input scalars"""
from tomviz import utils
import numpy as np
scalars = utils.get_scalars(dataset)
if scalars is None:
raise RuntimeError("No scalars found!")
# transform the dataset
result = np.sqrt(scalars)
# set the result as the new scalars.
utils.set_scalars(dataset, result)
def transform_scalars(dataset):
"""Define this method for Python operators that
transform input scalars"""
from tomviz import utils
import numpy as np
scalars = utils.get_scalars(dataset)
if scalars is None:
raise RuntimeError("No scalars found!")
if scalars.min() < 0:
print("WARNING: Square root of negative values results in NaN!")
else:
# transform the dataset
result = np.sqrt(scalars)
# set the result as the new scalars.
utils.set_scalars(dataset, result)
def transform_scalars(dataset):
"""Define this method for Python operators that
transform input scalars"""
from tomviz import utils
import numpy as np
scalars = utils.get_scalars(dataset)
if scalars is None:
raise RuntimeError("No scalars found!")
if scalars.min() < 0:
print("WARNING: Square root of negative values results in NaN!")
else:
# transform the dataset
result = np.sqrt(scalars)
# set the result as the new scalars.
utils.set_scalars(dataset, result)
def transform_scalars(self, dataset):
from tomviz import utils
import numpy as np
self.progress.maximum = NUMBER_OF_CHUNKS
scalars = utils.get_scalars(dataset)
if scalars is None:
raise RuntimeError("No scalars found!")
result = np.float32(scalars)
max = np.amax(scalars)
step = 0
for chunk in np.array_split(result, NUMBER_OF_CHUNKS):
if self.canceled:
return
chunk[:] = max - chunk
step += 1
self.progress.value = step
utils.set_scalars(dataset, result)
def transform_scalars(self, dataset):
from tomviz import utils
import numpy as np
self.progress.maximum = NUMBER_OF_CHUNKS
scalars = utils.get_scalars(dataset)
if scalars is None:
raise RuntimeError("No scalars found!")
result = np.float32(scalars)
max = np.amax(scalars)
step = 0
for chunk in np.array_split(result, NUMBER_OF_CHUNKS):
if self.canceled:
return
chunk[:] = max - chunk
step += 1
self.progress.value = step
utils.set_scalars(dataset, result)
def transform_scalars(self, dataset, label_value=1, principal_axis=0):
"""Computes the distance from the centroid of each connected component
in the label object with the given label_value to the given principal
axis and store that distance in each voxel of the label object connected
component. A principal_axis of 0 is first principal axis, 1 is the
second, and 2 is third.
"""
import numpy as np
from tomviz import itkutils
from tomviz import utils
self.progress.maximum = 100
self.progress.value = 0
STEP_PCT = [20, 60, 80, 100]
fd = dataset.GetFieldData()
axis_array = fd.GetArray('PrincipalAxes')
assert axis_array is not None, \
"Dataset does not have a PrincipalAxes field data array"
assert axis_array.GetNumberOfTuples() == 3, \
"PrincipalAxes array requires 3 tuples"
assert axis_array.GetNumberOfComponents() == 3, \
"PrincipalAxes array requires 3 components"
assert principal_axis >= 0 and principal_axis <= 2, \
"Invalid principal axis. Must be in range [0, 2]."
axis = np.array(axis_array.GetTuple(principal_axis))
center_array = fd.GetArray('Center')
assert center_array is not None, \
"Dataset does not have a Center field data array"
assert center_array.GetNumberOfTuples() == 1, \
"Center array requires 1 tuple"
assert center_array.GetNumberOfComponents() == 3, \
"Center array requires 3 components"
center = np.array(center_array.GetTuple(0))
# Blank out the undesired label values
scalars = utils.get_scalars(dataset)
scalars[scalars != label_value] = 0
utils.set_scalars(dataset, scalars)
self.progress.value = STEP_PCT[0]
# Get connected components of voxels labeled by label value
def connected_progress_func(fraction):
self.progress.value = \
int(fraction * (STEP_PCT[1] - STEP_PCT[0]) + STEP_PCT[0])
return self.canceled
utils.connected_components(dataset, 0, connected_progress_func)
# Get shape attributes
def label_progress_func(fraction):
self.progress.value = \
int(fraction * (STEP_PCT[2] - STEP_PCT[1]) + STEP_PCT[1])
return self.canceled
shape_label_map = \
itkutils.get_label_object_attributes(dataset, label_progress_func)
num_label_objects = shape_label_map.GetNumberOfLabelObjects()
# Map from label value to distance from principal axis. Used later to
# fill in distance array.
labels = utils.get_scalars(dataset)
max_label = np.max(labels)
label_value_to_distance = [0 for i in range(max_label + 1)]
for i in range(0, num_label_objects):
label_object = shape_label_map.GetNthLabelObject(i)
# Flip the centroid. I have verified that the x and z coordinates
# of the centroid coming out of the shape label objects are swapped,
# so I reverse it here.
centroid = np.flipud(np.array(label_object.GetCentroid()))
v = center - centroid
dot = np.dot(v, axis)
d = np.linalg.norm(v - dot*axis)
label_value_to_distance[label_object.GetLabel()] = d
distance = np.zeros(dataset.GetNumberOfPoints())
for i in range(len(labels)):
distance[i] = label_value_to_distance[labels[i]]
self.progress.value = STEP_PCT[3]
import vtk.util.numpy_support as np_s
distance_array = np_s.numpy_to_vtk(distance, deep=1)
distance_array.SetName('Distance')
dataset.GetPointData().SetScalars(distance_array)
def transform_scalars(self, dataset, label_value=1, principal_axis=0):
"""Computes the distance from the centroid of each connected component
in the label object with the given label_value to the given principal
axis and store that distance in each voxel of the label object connected
component. A principal_axis of 0 is first principal axis, 1 is the
second, and 2 is third.
"""
import numpy as np
from tomviz import itkutils
from tomviz import utils
self.progress.maximum = 100
self.progress.value = 0
STEP_PCT = [20, 60, 80, 100]
fd = dataset.GetFieldData()
axis_array = fd.GetArray('PrincipalAxes')
assert axis_array is not None, \
"Dataset does not have a PrincipalAxes field data array"
assert axis_array.GetNumberOfTuples() == 3, \
"PrincipalAxes array requires 3 tuples"
assert axis_array.GetNumberOfComponents() == 3, \
"PrincipalAxes array requires 3 components"
assert principal_axis >= 0 and principal_axis <= 2, \
"Invalid principal axis. Must be in range [0, 2]."
axis = np.array(axis_array.GetTuple(principal_axis))
center_array = fd.GetArray('Center')
assert center_array is not None, \
"Dataset does not have a Center field data array"
assert center_array.GetNumberOfTuples() == 1, \
"Center array requires 1 tuple"
assert center_array.GetNumberOfComponents() == 3, \
"Center array requires 3 components"
center = np.array(center_array.GetTuple(0))
# Blank out the undesired label values
scalars = utils.get_scalars(dataset)
scalars[scalars != label_value] = 0
utils.set_scalars(dataset, scalars)
self.progress.value = STEP_PCT[0]
# Get connected components of voxels labeled by label value
def connected_progress_func(fraction):
self.progress.value = \
int(fraction * (STEP_PCT[1] - STEP_PCT[0]) + STEP_PCT[0])
return self.canceled
utils.connected_components(dataset, 0, connected_progress_func)
# Get shape attributes
def label_progress_func(fraction):
self.progress.value = \
int(fraction * (STEP_PCT[2] - STEP_PCT[1]) + STEP_PCT[1])
return self.canceled
shape_label_map = \
itkutils.get_label_object_attributes(dataset, label_progress_func)
num_label_objects = shape_label_map.GetNumberOfLabelObjects()
# Map from label value to distance from principal axis. Used later to
# fill in distance array.
labels = utils.get_scalars(dataset)
max_label = np.max(labels)
label_value_to_distance = [0 for i in range(max_label + 1)]
for i in range(0, num_label_objects):
label_object = shape_label_map.GetNthLabelObject(i)
# Flip the centroid. I have verified that the x and z coordinates
# of the centroid coming out of the shape label objects are swapped,
# so I reverse it here.
centroid = np.flipud(np.array(label_object.GetCentroid()))
v = center - centroid
dot = np.dot(v, axis)
d = np.linalg.norm(v - dot * axis)
label_value_to_distance[label_object.GetLabel()] = d
distance = np.zeros(dataset.GetNumberOfPoints())
for i in range(len(labels)):
distance[i] = label_value_to_distance[labels[i]]
self.progress.value = STEP_PCT[3]
import vtk.util.numpy_support as np_s
distance_array = np_s.numpy_to_vtk(distance, deep=1)
distance_array.SetName('Distance')
dataset.GetPointData().SetScalars(distance_array)