def get_x(self, y):
""" tuple of x-values at intensity y
Return distance values based on interpolation of source data for a
supplied y value.
Parameters
----------
y : float
Returns
-------
tuple : (x1, x2, ...)
"""
dose_step = (max(self.y) - min(self.y)) / 100
x_ = self.resample_y(dose_step).x
y_ = self.resample_y(dose_step).y
dists = []
for i in range(1, len(x_)):
val = None
if (y_[i] - y) * (y_[i - 1] - y) < 0:
val = x_[i] - ((y_[i] - y) /
(y_[i] - y_[i - 1])) * (x_[i] - x_[i - 1])
elif np.isclose(y_[i], y):
val = x_[i]
if val and (val not in dists):
dists.append(val)
return tuple(dists)
def is_anonymised_dataset(ds, ignore_private_tags=False):
r"""Check whether a DICOM dataset has been (fully) anonymised.
This function specifically checks whether the dataset has been
anonymised using a PyMedPhys anonymiser. It is very likely that it
will return ``False`` for an anonymous dataset that was anonymised
using a different tool.
Parameters
----------
ds : ``pydicom.dataset.Dataset``
The DICOM dataset to check for anonymity
ignore_private_tags : ``bool``, optional
If set to ``False``, ``is_anonymised_dataset()`` will return
``False`` if any private (non-standard) DICOM tags exist in
``ds``. Set to ``True`` to ignore private tags when checking for
anonymity. Do so with caution, since private tags may contain
identifying information. Defaults to ``False``.
Returns
-------
is_anonymised : ``bool``
`True` if `ds` has been anonymised, `False` otherwise.
"""
for elem in ds:
if elem.keyword in get_default_identifying_keywords():
dummy_value = get_anonymous_replacement_value(elem.keyword)
if not elem.value in ("", [], dummy_value, None):
if elem.VR == "DS" and np.isclose(float(elem.value),
float(dummy_value)):
continue
logging.info("%s is not considered to be anonymised",
elem.name)
return False
elif elem.tag.is_private and not ignore_private_tags:
logging.info(
"%s is private and private tags are not being ignored",
elem.tag)
return False
elif elem.VR == "SQ":
contents_are_anonymous = True
for seq in elem.value:
contents_are_anonymous = is_anonymised_dataset(
seq, ignore_private_tags)
if not contents_are_anonymous:
logging.info(
"%s contained an element not considered to be anonymised",
elem.name,
)
return False
return True
def get_increment(self):
""" minimum step-size increment
Returns
-------
increment : float
"""
steps = np.diff(self.x)
if np.isclose(steps.min(), steps.mean()):
return steps.mean()
else:
return steps.min()