lsst.afw.cameraGeom._assembleImage.AmplifierIsolator Class Reference

Public Member Functions
def	__init__ (self, amplifier, parent_bbox, parent_detector)
def	subimage_bbox (self)
def	transform_subimage (self, subimage)
def	make_detector (self)
def	apply (cls, parent_exposure, amplifier)

Detailed Description

A class that can extracts single-amplifier subimages from trimmed or
untrimmed assembled images and transforms them to a particular orientation
and offset.

Callers who have a in-memory assembled `lsst.afw.image.Exposure` should
generally just use the `apply` class method.  Other methods can be used to
implement subimage loads of on on-disk images (e.g. formatter classes in
``obs_base``) or obtain subsets from other image classes.

Parameters
----------
amplifier : `Amplifier`
    Amplifier object that identifies the amplifier to load and sets the
    orientation and offset of the returned subimage.
parent_bbox : `lsst.geom.Box2I`
    Bounding box of the assembled parent image.  This must be equal to
    either ``parent_detector.getBBox()`` or
    ``parent_detector.getRawBBox()``; which one is used to determine
    whether the parent image (and hence the amplifier subimages) is
    trimmed.
parent_detector : `Detector`
    Detector object that describes the parent image.

Definition at line 131 of file _assembleImage.py.

Constructor & Destructor Documentation

__init__()

def lsst.afw.cameraGeom._assembleImage.AmplifierIsolator.__init__	(		self,
			amplifier,
			parent_bbox,
			parent_detector
	)

Definition at line 156 of file _assembleImage.py.

    def __init__(self, amplifier, parent_bbox, parent_detector):
        self._amplifier = amplifier
        self._parent_detector = parent_detector
        self._parent_amplifier = self._parent_detector[self._amplifier.getName()]
        self._is_parent_trimmed = (parent_bbox == self._parent_detector.getBBox())
        self._amplifier_comparison = self._amplifier.compareGeometry(self._parent_amplifier)
        if self._is_parent_trimmed:
            # We only care about the final bounding box; don't check e.g.
            # overscan regions for consistency.
            if self._parent_amplifier.getBBox() != self._amplifier.getBBox():
                raise ValueError(
                    f"The given amplifier's trimmed bounding box ({self._amplifier.getBBox()}) is not the "
                    "same as the trimmed bounding box of the same amplifier in the parent image "
                    f"({self._parent_amplifier.getBBox()})."
                )
        else:
            # Parent is untrimmed, so we need all regions to be consistent
            # between the amplifiers modulo flips and offsets.
            if self._amplifier_comparison & self._amplifier_comparison.REGIONS_DIFFER:
                raise ValueError(
                    "The given amplifier's subregions are fundamentally incompatible with those of the "
                    "parent image's amplifier."
                )
 

Member Function Documentation

apply()

def lsst.afw.cameraGeom._assembleImage.AmplifierIsolator.apply	(		cls,
			parent_exposure,
			amplifier
	)

Obtain a single-amplifier `lsst.afw.image.Exposure` subimage that
masquerades as full-detector image for a single-amp detector.

Parameters
----------
parent_exposure : `lsst.afw.image.Exposure`
    Parent image to obtain a subset from.
    `~lsst.afw.image.Exposure.getDetector` must not return `None`.
amplifier : `Amplifier`
    Target amplifier for the subimage.  May differ from the amplifier
    obtained by ``parent_exposure.getDetector()[amplifier.getName()]``
    only by flips and differences in `~Amplifier.getRawXYOffset`.

Returns
-------
subimage : `lsst.afw.image.Exposure`
    Exposure subimage for the target amplifier, with the
    orientation and XY0 described by that amplifier, and a single-amp
    detector holding a copy of that amplifier.

Notes
-----
Because we use the target amplifier's bounding box as the bounding box
of the detector attached to the returned exposure, other exposure
components that are passed through unmodified (e.g. the WCS) should
still be valid for the single-amp exposure after it is trimmed and
"assembled".  Unlike most trimmed+assembled images, however, it will
have a nonzero XY0, and code that (incorrectly!) does not pay attention
to XY0 may break.

Definition at line 249 of file _assembleImage.py.

    def apply(cls, parent_exposure, amplifier):
        """Obtain a single-amplifier `lsst.afw.image.Exposure` subimage that
        masquerades as full-detector image for a single-amp detector.
 
        Parameters
        ----------
        parent_exposure : `lsst.afw.image.Exposure`
            Parent image to obtain a subset from.
            `~lsst.afw.image.Exposure.getDetector` must not return `None`.
        amplifier : `Amplifier`
            Target amplifier for the subimage.  May differ from the amplifier
            obtained by ``parent_exposure.getDetector()[amplifier.getName()]``
            only by flips and differences in `~Amplifier.getRawXYOffset`.
 
        Returns
        -------
        subimage : `lsst.afw.image.Exposure`
            Exposure subimage for the target amplifier, with the
            orientation and XY0 described by that amplifier, and a single-amp
            detector holding a copy of that amplifier.
 
        Notes
        -----
        Because we use the target amplifier's bounding box as the bounding box
        of the detector attached to the returned exposure, other exposure
        components that are passed through unmodified (e.g. the WCS) should
        still be valid for the single-amp exposure after it is trimmed and
        "assembled".  Unlike most trimmed+assembled images, however, it will
        have a nonzero XY0, and code that (incorrectly!) does not pay attention
        to XY0 may break.
        """
        instance = cls(amplifier, parent_bbox=parent_exposure.getBBox(),
                       parent_detector=parent_exposure.getDetector())
        result = instance.transform_subimage(parent_exposure[instance.subimage_bbox])
        result.setDetector(instance.make_detector())
        return result

make_detector()

def lsst.afw.cameraGeom._assembleImage.AmplifierIsolator.make_detector

(

self

)

Create a single-amplifier detector that describes the transformed
subimage.

Returns
-------
detector : `Detector`
    Detector object with a single amplifier, a trimmed bounding box
    equal to the amplifier's trimmed bounding box, and no crosstalk.

Definition at line 231 of file _assembleImage.py.

    def make_detector(self):
        """Create a single-amplifier detector that describes the transformed
        subimage.
 
        Returns
        -------
        detector : `Detector`
            Detector object with a single amplifier, a trimmed bounding box
            equal to the amplifier's trimmed bounding box, and no crosstalk.
        """
        detector = self._parent_detector.rebuild()
        detector.clear()
        detector.append(self._amplifier.rebuild())
        detector.setBBox(self._amplifier.getBBox())
        detector.unsetCrosstalk()
        return detector.finish()
 

subimage_bbox()

def lsst.afw.cameraGeom._assembleImage.AmplifierIsolator.subimage_bbox

(

self

)

The bounding box of the target amplifier in the parent image
(`lsst.geom.Box2I`).

Definition at line 181 of file _assembleImage.py.

    def subimage_bbox(self):
        """The bounding box of the target amplifier in the parent image
        (`lsst.geom.Box2I`).
        """
        if self._is_parent_trimmed:
            return self._parent_amplifier.getBBox()
        else:
            return self._parent_amplifier.getRawBBox()
 

transform_subimage()

def lsst.afw.cameraGeom._assembleImage.AmplifierIsolator.transform_subimage	(		self,
			subimage
	)

Transform an already-extracted subimage to match the orientation
and offset of the target amplifier.

Parameters
----------
subimage : image-like
    The subimage to transform; may be any of `lsst.afw.image.Image`,
    `lsst.afw.image.Mask`, `lsst.afw.image.MaskedImage`, and
    `lsst.afw.image.Exposure`.

Returns
-------
transformed : image-like
    Transformed image of the same type as ``subimage``.

Definition at line 190 of file _assembleImage.py.

    def transform_subimage(self, subimage):
        """Transform an already-extracted subimage to match the orientation
        and offset of the target amplifier.
 
        Parameters
        ----------
        subimage : image-like
            The subimage to transform; may be any of `lsst.afw.image.Image`,
            `lsst.afw.image.Mask`, `lsst.afw.image.MaskedImage`, and
            `lsst.afw.image.Exposure`.
 
        Returns
        -------
        transformed : image-like
            Transformed image of the same type as ``subimage``.
        """
        from lsst.afw.math import flipImage
        if hasattr(subimage, "getMaskedImage"):
            # flipImage doesn't support Exposure natively.
            # And sadly, there's no way to write to an existing MaskedImage,
            # so we need to make yet another copy.
            result = subimage.clone()
            result.setMaskedImage(
                flipImage(
                    subimage.getMaskedImage(),
                    self._amplifier_comparison & self._amplifier_comparison.FLIPPED_X,
                    self._amplifier_comparison & self._amplifier_comparison.FLIPPED_Y,
                )
            )
        else:
            result = flipImage(
                subimage,
                self._amplifier_comparison & self._amplifier_comparison.FLIPPED_X,
                self._amplifier_comparison & self._amplifier_comparison.FLIPPED_Y,
            )
        if self._is_parent_trimmed:
            result.setXY0(self._amplifier.getBBox().getMin())
        else:
            result.setXY0(self._amplifier.getRawBBox().getMin() + self._amplifier.getRawXYOffset())
        return result
 

---

The documentation for this class was generated from the following file:

• /j/snowflake/release/lsstsw/stack/lsst-scipipe-0.7.0/Linux64/afw/22.0.1-42-gca6935d93+ba5e5ca3eb/python/lsst/afw/cameraGeom/_assembleImage.py