API

class donuts.Donuts(refimage, image_ext=0, exposure='EXPTIME', normalise=True, subtract_bkg=True, prescan_width=0, overscan_width=0, scan_direction='x', border=64, ntiles=32, image_class=<class 'donuts.image.Image'>)

This class provides methods for measuring shifts between a series of images of the same star field. First we initialise the object and generate a reference image. Subsequent images are aligned to this frame of this reference image.

Attributes

None

Initialise and generate a reference image. This reference image is used for measuring frame to frame offsets.

Parameters:

refimage : str

The image representing the reference frame.

image_ext: int, optional

The fits image extension to extract. The default is 0.

exposure : str, optional

Fits header keyword for exposure time. The default is EXPTIME.

normalise : bool, optional

Convert image counts to counts/s. The default is True.

subtract_bkg : bool, optional

Subtract the sky background. The default is True.

prescan_width : int, optional

Width of prescan region (left) in pixels. The default is 0.

overscan_width : int, optional

Width of overscan region (right) in pixels. The default is 0.

scan_direction : str, optional

Direction along which the pre/overscan regions are found (‘x’ | ‘y’)

border : int, optional

Width of exclusion area to avoid errors from CCD edge effects. The default is 64.

ntiles : int, optional

Number of tiles used to sample the sky background. The default is 32.

Returns:

None

Raises:

None

construct_object(filename)

Builds an image_class instance which performs most of the work. See the Image class for more information.

Parameters:

filename : str

FITS file to open and build an image_class instance from.

Returns:

image_class instance

Raises:

None

measure_shift(checkimage_filename)

Generate a check image and measure its offset from the reference This is done using the same settings as the reference image.

Parameters:

checkimage_filename : str

Image filename to compare to reference image

Returns:

image: Image

Instance of an Image object, which has the x and y atributes storing the value of the shift between the chosen image and reference image (passed to the Donuts constructor)

Raises:

None

print_summary()

Print a summary of the current settings

Parameters:None Returns:None Raises:None

class donuts.image.Image(data, header=None)

Low level class which handles the image transformations and cross correlation with another Image class.

compute_offset(reference_image)

Given another Image object, compute the shift in pixel units.

This method sets self.x and self.y to the pixel shift, and returns the instance so in effect the user gets a “new” Image instance with these variables set.

Parameters:

reference_image : Image

The reference image to compare to. Typically when called using the Donuts class this will be whatever was defined as the “reference” image

Returns

——-

self : Image

The current Image instance

Raises:

None

compute_projections()

Compute the projection profiles. Follows the following logic:

• if the image has not been trimmed and not background subtracted, then use the raw pixels
• otherwise if the image has been trimmed then use the trimmed pixels
• otherwise use the background subtracted pixels

See _projection_from_image() for details of the projection calculation.

Parameters:

None

Returns:

self : Image

The current Image instance

Raises:

None

normalise(exposure_keyword='EXPOSURE')

Convert the image data into ADU per second

Parameters:

exposure_keyword : str

Fits header keyword for exposure time. The default is EXPTIME.

Returns:

self : Image

The current Image instance

Raises:

RuntimeError if the image region has not been trimmed (see the ``trim``

method).

postconstruct_hook()

Hook to modify the class after any standard processing

To add functionality, alter backsub_region()

preconstruct_hook()

Hook to modify the class before any standard processing

To add functionality, alter raw_image()

remove_background(ntiles=32)

Subtract the background from the image. See _generate_bkg_map() for more details

Parameters:

ntiles : int

Number of tiles used to sample the sky background. The default is 32.

Returns:

self : Image

The current Image instance

Raises:

None

trim(prescan_width=0, overscan_width=0, scan_direction='x', border=64)

Remove the optional prescan and overscan from the image, as well as the outer n rows/colums of the image. Finally ensure the imaging region is the correct dimensions for skimage.transform.resize() (i.e. a multiple of 16.)

Parameters:

prescan_width : int

Remove the first prescan_width columns from the image, assuming the are not in the imaging region.

overscan_width : int

Remove the last overscan_width columns from the image, assuming the are not in the imaging region.

scan_direction : ‘x’ | ‘y’

Direction along which the pre/overscans occur. If along left and right side, select ‘x’. If along the top and bottom of the image select ‘y’

border : int

Ignore the first/last border rows/columns from the image, assuming that they are not “typical”, a common case with edge effects in CCDs.

Returns:

self : Image

The current Image instance

Raises:

None