# Licensed under a 3-clause BSD style license - see LICENSE.rst
"""
Base aperture classes.
"""
import abc
import inspect
import warnings
from copy import deepcopy
import numpy as np
from astropy.coordinates import SkyCoord
from astropy.utils import lazyproperty
from photutils.aperture.bounding_box import BoundingBox
from photutils.aperture.mask import ApertureMask
from photutils.geometry._batch_photometry import batch_aperture_sums
from photutils.utils._deprecation import deprecated_positional_kwargs
__all__ = ['Aperture', 'PixelAperture', 'SkyAperture']
[docs]
class Aperture(metaclass=abc.ABCMeta):
"""
Abstract base class for all apertures.
"""
_params = ()
def __len__(self):
if self.isscalar:
msg = f'A scalar {self.__class__.__name__!r} object has no len()'
raise TypeError(msg)
return self.shape[0]
def __getitem__(self, index):
if self.isscalar:
msg = (f'A scalar {self.__class__.__name__!r} object cannot be '
'indexed')
raise TypeError(msg)
kwargs = {}
for param in self._params:
if param == 'positions':
# Slice the positions array
kwargs[param] = getattr(self, param)[index]
else:
kwargs[param] = getattr(self, param)
return self.__class__(**kwargs)
def __iter__(self):
for i in range(len(self)):
yield self.__getitem__(i)
def _positions_str(self, *, prefix=None):
if isinstance(self, PixelAperture):
return np.array2string(self.positions, separator=', ',
prefix=prefix)
if isinstance(self, SkyAperture):
return repr(self.positions)
msg = 'Aperture must be a subclass of PixelAperture or SkyAperture'
raise TypeError(msg)
def __repr__(self):
prefix = f'{self.__class__.__name__}'
cls_info = []
for param in self._params:
if param == 'positions':
cls_info.append(self._positions_str(prefix=prefix))
else:
cls_info.append(f'{param}={getattr(self, param)}')
cls_info = ', '.join(cls_info)
return f'<{prefix}({cls_info})>'
def __str__(self):
cls_info = [('Aperture', self.__class__.__name__)]
for param in self._params:
if param == 'positions':
prefix = 'positions'
cls_info.append((prefix,
self._positions_str(prefix=prefix + ': ')))
else:
cls_info.append((param, getattr(self, param)))
fmt = [f'{key}: {val}' for key, val in cls_info]
return '\n'.join(fmt)
def __eq__(self, other):
"""
Equality operator for `Aperture`.
All Aperture properties are compared for strict equality except
for Quantity parameters, which allow for different units if they
are directly convertible.
"""
if not isinstance(other, self.__class__):
return False
self_params = list(self._params)
other_params = list(other._params)
# Check that both have identical parameters
if self_params != other_params:
return False
# Now check the parameter values.
# Note that Quantity comparisons allow for different units if they
# are directly convertible (e.g., 1.0 * u.deg == 60.0 * u.arcmin)
try:
for param in self_params:
# np.any is used for SkyCoord array comparisons
if np.any(getattr(self, param) != getattr(other, param)):
return False
except TypeError:
# TypeError is raised from SkyCoord comparison when they do
# not have equivalent frames. Here return False instead of
# the TypeError.
return False
return True
def __ne__(self, other):
"""
Inequality operator for `Aperture`.
"""
return not self == other
@property
def _lazyproperties(self):
"""
A list of all class lazyproperties (even in superclasses).
"""
def islazyproperty(obj):
return isinstance(obj, lazyproperty)
return [i[0] for i in inspect.getmembers(self.__class__,
predicate=islazyproperty)]
[docs]
def copy(self):
"""
Make a deep copy of this object.
Returns
-------
result : `Aperture`
A deep copy of the Aperture object.
"""
params_copy = {}
for param in list(self._params):
params_copy[param] = deepcopy(getattr(self, param))
return self.__class__(**params_copy)
[docs]
@abc.abstractmethod
def positions(self):
"""
The aperture positions, as an array of (x, y) coordinates or a
`~astropy.coordinates.SkyCoord`.
"""
@lazyproperty
def shape(self):
"""
The shape of the instance.
"""
if isinstance(self.positions, SkyCoord):
return self.positions.shape
return self.positions.shape[:-1]
@lazyproperty
def isscalar(self):
"""
Whether the instance is scalar (i.e., a single position).
"""
return self.shape == ()
[docs]
class PixelAperture(Aperture):
"""
Abstract base class for apertures defined in pixel coordinates.
"""
@lazyproperty
def _default_patch_properties(self):
"""
A dictionary of default matplotlib.patches.Patch properties.
"""
mpl_params = {}
# matplotlib.patches.Patch default is ``fill=True``
mpl_params['fill'] = False
return mpl_params
@staticmethod
def _translate_mask_method(method, subpixels):
"""
Translate the mask method and subpixels parameters to the values
used by the low-level `photutils.geometry` functions.
Parameters
----------
method : {'exact', 'center', 'subpixel'}
The mask method.
subpixels : int
The number of subpixels for the 'subpixel' method.
Returns
-------
use_exact : int
Whether to use exact method (1) or not (0).
subpixels : int
The number of subpixels for subpixel method.
"""
if method not in ('center', 'subpixel', 'exact'):
msg = f'Invalid mask method: {method}'
raise ValueError(msg)
if ((method == 'subpixel')
and (not isinstance(subpixels, int) or subpixels <= 0)):
msg = 'subpixels must be a strictly positive integer'
raise ValueError(msg)
if method == 'center':
use_exact = 0
subpixels = 1
elif method == 'subpixel':
use_exact = 0
elif method == 'exact':
use_exact = 1
subpixels = 1
return use_exact, subpixels
@property
@abc.abstractmethod
def _xy_extents(self):
"""
The (x, y) extents of the aperture measured from the center
position.
In other words, the (x, y) extents are half of the aperture
minimal bounding box size in each dimension.
"""
@lazyproperty
def _positions(self):
"""
The aperture positions, always as a 2D ndarray.
"""
return np.atleast_2d(self.positions)
@lazyproperty
def _bbox(self):
"""
The minimal bounding box for the aperture, always as a list of
`~photutils.aperture.BoundingBox` instances.
"""
x_delta, y_delta = self._xy_extents
xmin = self._positions[:, 0] - x_delta
xmax = self._positions[:, 0] + x_delta
ymin = self._positions[:, 1] - y_delta
ymax = self._positions[:, 1] + y_delta
return [BoundingBox.from_float(x0, x1, y0, y1)
for x0, x1, y0, y1 in zip(xmin, xmax, ymin, ymax, strict=True)]
@lazyproperty
def bbox(self):
"""
The minimal bounding box for the aperture.
If the aperture is scalar then a single
`~photutils.aperture.BoundingBox` is returned, otherwise a list
of `~photutils.aperture.BoundingBox` is returned.
"""
if self.isscalar:
return self._bbox[0]
return self._bbox
@lazyproperty
def _centered_edges(self):
"""
A list of ``(xmin, xmax, ymin, ymax)`` tuples, one for each
position, of the pixel edges after recentering the aperture at
the origin.
These pixel edges are used by the low-level `photutils.geometry`
functions.
"""
edges = []
for position, bbox in zip(self._positions, self._bbox, strict=True):
xmin = bbox.ixmin - 0.5 - position[0]
xmax = bbox.ixmax - 0.5 - position[0]
ymin = bbox.iymin - 0.5 - position[1]
ymax = bbox.iymax - 0.5 - position[1]
edges.append((xmin, xmax, ymin, ymax))
return edges
@property
@abc.abstractmethod
def area(self):
"""
The exact geometric area of the aperture shape.
Use the `area_overlap` method to return the area of overlap
between the data and the aperture, taking into account the
aperture mask method, masked data pixels (``mask`` keyword), and
partial/no overlap of the aperture with the data.
Returns
-------
area : float
The aperture area.
See Also
--------
area_overlap
"""
[docs]
def area_overlap(self, data, *, mask=None, method='exact', subpixels=5):
"""
Return the area of overlap between the data and the aperture.
This method takes into account the aperture mask method, masked
data pixels (``mask`` keyword), and partial/no overlap of the
aperture with the data. In other words, it returns the area that
used to compute the aperture sum (assuming identical inputs).
Use the `area` method to calculate the exact analytical area of
the aperture shape.
Parameters
----------
data : array_like or `~astropy.units.Quantity`
A 2D array.
mask : array_like (bool), optional
A boolean mask with the same shape as ``data`` where a
`True` value indicates the corresponding element of ``data``
is masked. Masked data are excluded from the area overlap.
method : {'exact', 'center', 'subpixel'}, optional
The method used to determine the pixel weights (the fraction
of the pixel area covered by the aperture):
* ``'exact'`` (default):
Calculates the exact geometric overlap area. Weights are
continuous in the range [0, 1].
* ``'center'``:
Binary weighting based on the pixel center. Weights are
either 0 or 1.
* ``'subpixel'``:
Approximates the overlap by averaging binary samples on a
subgrid. The number of samples is set by the ``subpixels``
parameter. Weights are discrete in the range [0, 1].
subpixels : int, optional
The subsampling factor per axis used when
``method='subpixel'``. Each pixel is divided into a grid of
``subpixels**2`` subpixels to approximate the overlap. This
parameter is ignored for other methods.
Returns
-------
areas : float or array_like
The area (in pixels**2) of overlap between the data and the
aperture.
See Also
--------
area
"""
apermasks = self.to_mask(method=method, subpixels=subpixels)
if self.isscalar:
apermasks = (apermasks,)
if mask is not None:
mask = np.asarray(mask)
if mask.shape != data.shape:
msg = 'mask and data must have the same shape'
raise ValueError(msg)
areas = []
for apermask in apermasks:
slc_large, slc_small = apermask.get_overlap_slices(data.shape)
# If the aperture does not overlap the data, return np.nan
if slc_large is None:
area = np.nan
else:
aper_weights = apermask.data[slc_small]
if mask is not None:
aper_weights[mask[slc_large]] = 0.0
area = np.sum(aper_weights)
areas.append(area)
areas = np.array(areas)
if self.isscalar:
return areas[0]
return areas
[docs]
@deprecated_positional_kwargs(since='3.0', until='4.0')
def to_mask(self, method='exact', subpixels=5):
"""
Return a mask for the aperture.
Parameters
----------
method : {'exact', 'center', 'subpixel'}, optional
The method used to determine the pixel weights (the fraction
of the pixel area covered by the aperture):
* ``'exact'`` (default):
Calculates the exact geometric overlap area. Weights are
continuous in the range [0, 1].
* ``'center'``:
Binary weighting based on the pixel center. Weights are
either 0 or 1.
* ``'subpixel'``:
Approximates the overlap by averaging binary samples on a
subgrid. The number of samples is set by the ``subpixels``
parameter. Weights are discrete in the range [0, 1].
subpixels : int, optional
The subsampling factor per axis used when
``method='subpixel'``. Each pixel is divided into a grid of
``subpixels**2`` subpixels to approximate the overlap. This
parameter is ignored for other methods.
Returns
-------
mask : `~photutils.aperture.ApertureMask` or list of \
`~photutils.aperture.ApertureMask`
A mask for the aperture. If the aperture is scalar then
a single `~photutils.aperture.ApertureMask` is returned,
otherwise a list of `~photutils.aperture.ApertureMask` is
returned.
"""
use_exact, subpixels = self._translate_mask_method(method, subpixels)
masks = []
for bbox, edges in zip(self._bbox, self._centered_edges, strict=True):
ny, nx = bbox.shape
overlap = self._compute_overlap(
edges, nx, ny, use_exact, subpixels)
masks.append(ApertureMask(overlap, bbox))
if self.isscalar:
return masks[0]
return masks
@abc.abstractmethod
def _compute_overlap(self, edges, nx, ny, use_exact, subpixels):
"""
Compute the overlap of the aperture for a single position.
Parameters
----------
edges : tuple of float
The ``(xmin, xmax, ymin, ymax)`` pixel edges centered at
the origin.
nx, ny : int
The number of pixels in x and y.
use_exact : int
Whether to use exact method (1) or not (0).
subpixels : int
The number of subpixels for subpixel method.
Returns
-------
overlap : 2D `~numpy.ndarray`
The overlap array.
"""
def _do_mask_photometry(self, data, *, error, mask, method, subpixels):
"""
Perform aperture photometry using per-source aperture masks.
This is the fallback code path for apertures or inputs that are
not supported by the batch Cython driver.
Parameters
----------
data : `~numpy.ndarray`
The 2D array on which to perform photometry, with any units
already stripped.
error, mask, method, subpixels
See `do_photometry`. Any units must already be stripped from
``error``.
Returns
-------
aperture_sums, aperture_sum_errs : `~numpy.ndarray`
The aperture sums and errors.
"""
apermasks = self.to_mask(method=method, subpixels=subpixels)
if self.isscalar:
apermasks = (apermasks,)
aperture_sums = []
aperture_sum_errs = []
with warnings.catch_warnings():
# Ignore multiplication with non-finite data values
warnings.simplefilter('ignore', RuntimeWarning)
for apermask in apermasks:
(slc_large,
aper_weights,
pixel_mask) = apermask._get_overlap_cutouts(data.shape,
mask=mask)
# No overlap of the aperture with the data
if slc_large is None:
aperture_sums.append(np.nan)
aperture_sum_errs.append(np.nan)
continue
values = (data[slc_large] * aper_weights)[pixel_mask]
aperture_sums.append(values.sum())
if error is not None:
variance = (error[slc_large]**2
* aper_weights)[pixel_mask]
aperture_sum_errs.append(np.sqrt(variance.sum()))
return np.array(aperture_sums), np.array(aperture_sum_errs)
def _batch_shape_params(self):
"""
The aperture shape code and parameters for the batch Cython
photometry driver.
Returns
-------
spec : tuple or `None`
A ``(shape_code, params)`` tuple, where
``shape_code`` is one of the shape codes defined in
`photutils.geometry._batch_photometry` and ``params``
is a tuple of the aperture shape parameters expected by
`~photutils.geometry._batch_photometry.batch_aperture_sums`
for that shape. `None` is returned if batch photometry is
not supported for this aperture, in which case the slower
mask-based code path is used.
Notes
-----
The batch driver is used only if this hook is defined in the
aperture instance's own class (see `_do_batch_photometry`),
so subclasses must define this method (e.g., by calling
``super()``) to opt in to the batch code path.
"""
return
def _do_batch_photometry(self, data, *, error, mask, method, subpixels):
"""
Perform aperture photometry using the batch Cython driver.
The batch driver computes results identical to the mask-based
code path, but without creating per-source mask arrays or making
per-source Python calls.
Parameters
----------
data : `~numpy.ndarray`
The 2D array on which to perform photometry, with any units
already stripped.
error, mask, method, subpixels
See `do_photometry`. Any units must already be stripped from
``error``.
Returns
-------
result : tuple of `~numpy.ndarray` or `None`
A ``(aperture_sums, aperture_sum_errs)`` tuple of float64
arrays, or `None` if the batch driver does not support this
aperture or these inputs (in which case the caller should
use the mask-based code path).
"""
# Use the batch driver only if the aperture's own class defines
# the _batch_shape_params hook. Subclasses that do not define
# it may override other behavior (e.g., to_mask) that the batch
# driver would not honor, so they use the mask-based code path.
if '_batch_shape_params' not in type(self).__dict__:
return None
spec = self._batch_shape_params()
if spec is None:
return None
def _supported(arr):
return (type(arr) is np.ndarray and arr.dtype.kind in 'fiub'
and arr.dtype.itemsize <= 8)
if not _supported(data) or (error is not None
and not _supported(error)):
return None
if mask is not None:
if (not isinstance(mask, np.ndarray) or mask.dtype != bool
or mask.shape != data.shape):
return None
mask = np.ascontiguousarray(mask, dtype=np.uint8)
use_exact, subpixels = self._translate_mask_method(method, subpixels)
shape_code, params = spec
if error is not None:
error = np.ascontiguousarray(error, dtype=np.float64)
ext_x, ext_y = self._xy_extents
sums, errs, overlap = batch_aperture_sums(
np.ascontiguousarray(data, dtype=np.float64), error, mask,
np.ascontiguousarray(self._positions, dtype=np.float64),
shape_code, np.array(params, dtype=np.float64),
float(ext_x), float(ext_y), use_exact, subpixels)
if error is None:
# Match the mask-based path, which collects one NaN per
# non-overlapping source when error is not input.
errs = np.full(np.count_nonzero(~overlap), np.nan)
return sums, errs
[docs]
@deprecated_positional_kwargs(since='3.0', until='4.0')
def do_photometry(self, data, error=None, mask=None, method='exact',
subpixels=5):
"""
Perform aperture photometry on the input data.
Parameters
----------
data : array_like or `~astropy.units.Quantity` instance
The 2D array on which to perform photometry. ``data`` should
be background subtracted.
error : array_like or `~astropy.units.Quantity`, optional
The pixel-wise Gaussian 1-sigma errors of the input
``data``. ``error`` is assumed to include *all* sources
of error, including the Poisson error of the sources (see
`~photutils.utils.calc_total_error`). ``error`` must have
the same shape as the input ``data``.
mask : array_like (bool), optional
A boolean mask with the same shape as ``data`` where a
`True` value indicates the corresponding element of ``data``
is masked. Masked data are excluded from all calculations.
method : {'exact', 'center', 'subpixel'}, optional
The method used to determine the pixel weights (the fraction
of the pixel area covered by the aperture):
* ``'exact'`` (default):
Calculates the exact geometric overlap area. Weights are
continuous in the range [0, 1].
* ``'center'``:
Binary weighting based on the pixel center. Weights are
either 0 or 1.
* ``'subpixel'``:
Approximates the overlap by averaging binary samples on a
subgrid. The number of samples is set by the ``subpixels``
parameter. Weights are discrete in the range [0, 1].
subpixels : int, optional
The subsampling factor per axis used when
``method='subpixel'``. Each pixel is divided into a grid of
``subpixels**2`` subpixels to approximate the overlap. This
parameter is ignored for other methods.
Returns
-------
aperture_sums : `~numpy.ndarray` or `~astropy.units.Quantity`
The sum within each aperture. The values are always
float64, regardless of the input ``data`` dtype.
aperture_sum_errs : `~numpy.ndarray` or `~astropy.units.Quantity`
The errors on the sum within each aperture. The values are
always float64, regardless of the input ``error`` dtype.
"""
data = np.asanyarray(data)
if data.ndim != 2:
msg = 'data must be a 2D array'
raise ValueError(msg)
if error is not None:
error = np.asanyarray(error)
if error.shape != data.shape:
msg = 'error and data must have the same shape'
raise ValueError(msg)
# Check Quantity inputs
unit = {getattr(arr, 'unit', None) for arr in (data, error)
if arr is not None}
if len(unit) > 1:
msg = ('If data or error has units, then they both must have '
'the same units')
raise ValueError(msg)
# Strip data and error units for performance
unit = unit.pop()
if unit is not None:
unit = data.unit
data = data.value
if error is not None:
error = error.value
result = self._do_batch_photometry(data, error=error, mask=mask,
method=method,
subpixels=subpixels)
if result is not None:
aperture_sums, aperture_sum_errs = result
else:
aperture_sums, aperture_sum_errs = self._do_mask_photometry(
data, error=error, mask=mask, method=method,
subpixels=subpixels)
# Apply units
if unit is not None:
aperture_sums <<= unit
aperture_sum_errs <<= unit
return aperture_sums, aperture_sum_errs
@staticmethod
def _make_annulus_path(patch_inner, patch_outer):
"""
Define a matplotlib annulus path from two patches.
This preserves the cubic Bézier curves (CURVE4) of the aperture
paths.
"""
import matplotlib.path as mpath
path_inner = patch_inner.get_path()
transform_inner = patch_inner.get_transform()
path_inner = transform_inner.transform_path(path_inner)
path_outer = patch_outer.get_path()
transform_outer = patch_outer.get_transform()
path_outer = transform_outer.transform_path(path_outer)
verts_inner = path_inner.vertices[:-1][::-1]
verts_inner = np.concatenate((verts_inner, [verts_inner[-1]]))
verts = np.vstack((path_outer.vertices, verts_inner))
codes = np.hstack((path_outer.codes, path_inner.codes))
return mpath.Path(verts, codes)
def _define_patch_params(self, *, origin=(0, 0), **kwargs):
"""
Define the aperture patch position and set any default
matplotlib patch keywords (e.g., ``fill=False``).
Parameters
----------
origin : array_like, optional
The ``(x, y)`` position of the origin of the displayed
image.
**kwargs : dict, optional
Any keyword arguments accepted by
`matplotlib.patches.Patch`.
Returns
-------
xy_positions : `~numpy.ndarray`
The aperture patch positions.
patch_params : dict
Any keyword arguments accepted by
`matplotlib.patches.Patch`.
"""
xy_positions = deepcopy(self._positions)
xy_positions[:, 0] -= origin[0]
xy_positions[:, 1] -= origin[1]
patch_params = self._default_patch_properties.copy()
patch_params.update(kwargs)
return xy_positions, patch_params
@abc.abstractmethod
def _to_patch(self, *, origin=(0, 0), **kwargs):
"""
Return a `~matplotlib.patches.Patch` for the aperture.
Parameters
----------
origin : array_like, optional
The ``(x, y)`` position of the origin of the displayed
image.
**kwargs : dict, optional
Any keyword arguments accepted by
`matplotlib.patches.Patch`.
Returns
-------
patch : `~matplotlib.patches.Patch` or list of \
`~matplotlib.patches.Patch`
A patch for the aperture. If the aperture is scalar then a
single `~matplotlib.patches.Patch` is returned, otherwise a
list of `~matplotlib.patches.Patch` is returned.
"""
[docs]
@deprecated_positional_kwargs(since='3.0', until='4.0')
def plot(self, ax=None, origin=(0, 0), **kwargs):
"""
Plot the aperture on a matplotlib `~matplotlib.axes.Axes`
instance.
Parameters
----------
ax : `matplotlib.axes.Axes` or `None`, optional
The matplotlib axes on which to plot. If `None`, then the
current `~matplotlib.axes.Axes` instance is used.
origin : array_like, optional
The ``(x, y)`` position of the origin of the displayed
image.
**kwargs : dict, optional
Any keyword arguments accepted by
`matplotlib.patches.Patch`.
Returns
-------
patch : list of `~matplotlib.patches.Patch`
A list of matplotlib patches for the plotted aperture. The
patches can be used, for example, when adding a plot legend.
"""
import matplotlib.pyplot as plt
if ax is None:
ax = plt.gca()
patches = self._to_patch(origin=origin, **kwargs)
if self.isscalar:
patches = (patches,)
for patch in patches:
ax.add_patch(patch)
return patches
[docs]
@abc.abstractmethod
def to_sky(self, wcs):
"""
Convert the aperture to a `SkyAperture` object defined in
celestial coordinates.
Parameters
----------
wcs : WCS object
A world coordinate system (WCS) transformation that
supports the `astropy shared interface for WCS
<https://docs.astropy.org/en/stable/wcs/wcsapi.html>`_
(e.g., `astropy.wcs.WCS`, `gwcs.wcs.WCS`).
Returns
-------
aperture : `SkyAperture` object
A `SkyAperture` object.
"""
[docs]
class SkyAperture(Aperture):
"""
Abstract base class for all apertures defined in celestial
coordinates.
"""
[docs]
@abc.abstractmethod
def to_pixel(self, wcs):
"""
Convert the aperture to a `PixelAperture` object defined in
pixel coordinates.
Parameters
----------
wcs : WCS object
A world coordinate system (WCS) transformation that
supports the `astropy shared interface for WCS
<https://docs.astropy.org/en/stable/wcs/wcsapi.html>`_
(e.g., `astropy.wcs.WCS`, `gwcs.wcs.WCS`).
Returns
-------
aperture : `PixelAperture` object
A `PixelAperture` object.
"""
def _aperture_metadata(aperture, *, index=''):
"""
Return a dictionary of aperture metadata.
Parameters
----------
aperture : `Aperture`
An aperture object.
index : str, optional
A string that will be prepended to each metadata key.
Returns
-------
meta : dict
A dictionary of aperture metadata
"""
params = aperture._params
meta = {}
meta[f'aperture{index}'] = aperture.__class__.__name__
for param in params:
if param != 'positions':
meta[f'aperture{index}_{param}'] = getattr(aperture, param)
return meta
