lsst.scarlet.lite.image Namespace Reference

Classes
class	Image
class	MismatchedBandsError
class	MismatchedBoxError

Functions
list[DTypeLike]	get_dtypes (*np.ndarray|Image|ScalarLike data)
DTypeLike	get_combined_dtype (*np.ndarray|Image|ScalarLike data)
Image	_operate_on_images (Image image1, Image|ScalarLike image2, Callable op)
Image	insert_image (Image main_image, Image sub_image, Callable op=operator.add)

Function Documentation

_operate_on_images()

Image lsst.scarlet.lite.image._operate_on_images ( Image image1, Image | ScalarLike image2, Callable op )

protected

Perform an operation on two images, that may or may not be spectrally
and spatially aligned.

Parameters
----------
image1:
    The image on the LHS of the operation
image2:
    The image on the RHS of the operation
op:
    The operation used to combine the images.

Returns
-------
image:
    The resulting combined image.

Definition at line 1234 of file image.py.

def _operate_on_images(image1: Image, image2: Image | ScalarLike, op: Callable) -> Image:
    """Perform an operation on two images, that may or may not be spectrally
    and spatially aligned.
 
    Parameters
    ----------
    image1:
        The image on the LHS of the operation
    image2:
        The image on the RHS of the operation
    op:
        The operation used to combine the images.
 
    Returns
    -------
    image:
        The resulting combined image.
    """
    if type(image2) in ScalarTypes:
        return image1.copy_with(data=op(image1.data, image2))
    image2 = cast(Image, image2)
    if image1.bands == image2.bands and image1.bbox == image2.bbox:
        # The images perfectly overlap, so just combine their results
        with np.errstate(divide="ignore", invalid="ignore"):
            result = op(image1.data, image2.data)
        return Image(result, bands=image1.bands, yx0=image1.yx0)
 
    if op != operator.add and op != operator.sub and image1.bands != image2.bands:
        msg = "Images with different bands can only be combined using addition and subtraction, "
        msg += f"got {op}, with bands {image1.bands}, {image2.bands}"
        raise ValueError(msg)
 
    # Use all of the bands in the first image
    bands = image1.bands
    # Add on any bands from the second image not contained in the first image
    bands = bands + tuple(band for band in image2.bands if band not in bands)
    # Create a box that contains both images
    bbox = image1.bbox | image2.bbox
    # Create an image that will contain both images
    if len(bands) > 0:
        shape = (len(bands),) + bbox.shape
    else:
        shape = bbox.shape
 
    if op == operator.add or op == operator.sub:
        dtype = get_combined_dtype(image1, image2)
        result = Image(np.zeros(shape, dtype=dtype), bands=bands, yx0=cast(tuple[int, int], bbox.origin))
        # Add the first image in place
        image1.insert_into(result, operator.add)
        # Use the operator to insert the second image
        image2.insert_into(result, op)
    else:
        # Project both images into the full bbox
        image1 = image1.project(bbox=bbox)
        image2 = image2.project(bbox=bbox)
        result = op(image1, image2)
    return result
 
 

get_combined_dtype()

DTypeLike lsst.scarlet.lite.image.get_combined_dtype

(

*np.ndarray | Image | ScalarLike

data

)

Get the combined dtype for a collection of arrays to prevent loss
of precision.

Parameters
----------
data:
    The arrays to use for calculating the dtype

Returns
-------
result: np.dtype
    The resulting dtype.

Definition at line 66 of file image.py.

def get_combined_dtype(*data: np.ndarray | Image | ScalarLike) -> DTypeLike:
    """Get the combined dtype for a collection of arrays to prevent loss
    of precision.
 
    Parameters
    ----------
    data:
        The arrays to use for calculating the dtype
 
    Returns
    -------
    result: np.dtype
        The resulting dtype.
    """
    dtypes = get_dtypes(*data)
    return max(dtypes)  # type: ignore
 
 

get_dtypes()

list[DTypeLike] lsst.scarlet.lite.image.get_dtypes

(

*np.ndarray | Image | ScalarLike

data

)

Get a list of dtypes from a list of arrays, images, or scalars

Parameters
----------
data:
    The arrays to use for calculating the dtype

Returns
-------
result:
    A list of datatypes.

Definition at line 44 of file image.py.

def get_dtypes(*data: np.ndarray | Image | ScalarLike) -> list[DTypeLike]:
    """Get a list of dtypes from a list of arrays, images, or scalars
 
    Parameters
    ----------
    data:
        The arrays to use for calculating the dtype
 
    Returns
    -------
    result:
        A list of datatypes.
    """
    dtypes: list[DTypeLike] = [None] * len(data)
    for d, element in enumerate(data):
        if hasattr(element, "dtype"):
            dtypes[d] = cast(np.ndarray, element).dtype
        else:
            dtypes[d] = np.dtype(type(element))
    return dtypes
 
 

insert_image()

Image lsst.scarlet.lite.image.insert_image	(	Image	main_image,
		Image	sub_image,
		Callable	op = operator.add )

Insert one image into another image

Parameters
----------
main_image:
    The image that will have `sub_image` insertd.
sub_image:
    The image that is inserted into `main_image`.
op:
    The operator to use for insertion
    (addition, subtraction, multiplication, etc.).

Returns
-------
main_image: Image
    The `main_image`, with the `sub_image` inserted in place.

Definition at line 1293 of file image.py.

) -> Image:
    """Insert one image into another image
 
    Parameters
    ----------
    main_image:
        The image that will have `sub_image` insertd.
    sub_image:
        The image that is inserted into `main_image`.
    op:
        The operator to use for insertion
        (addition, subtraction, multiplication, etc.).
 
    Returns
    -------
    main_image: Image
        The `main_image`, with the `sub_image` inserted in place.
    """
    if len(main_image.bands) == 0 and len(sub_image.bands) == 0:
        slices = sub_image.matched_slices(main_image.bbox)
        image_slices = slices[1]
        self_slices = slices[0]
    else:
        band_indices = sub_image.matched_spectral_indices(main_image)
        slices = sub_image.matched_slices(main_image.bbox)
        image_slices = (band_indices[0],) + slices[1]  # type: ignore
        self_slices = (band_indices[1],) + slices[0]  # type: ignore
 
    main_image._data[image_slices] = op(main_image.data[image_slices], sub_image.data[self_slices])
    return main_image