PSF Class

This class implements tools needed to work with the PSF. Below is an example of the Pandora PSF in pixel space. This class enables users to work with the PSF files and

• Evaluate it at any grid point (e.g. position, temperature, wavelength)
• Evaluate the PRF (i.e. the PSF integrated on a pixel grid)
• Find the gradient of the PSF or PRF

Class to work with abstract PSFs

PSF class for making PSFs, PRFs, and traces.

Parameters:

Name	Type	Description	Default
X	ndarray	Array of 1D vectors defining the value of each dimension for every element of psf_flux. Should have as many entries as psf_flux has dimensions.	required
psf_flux	ndarray	ND array of flux values	required
dimension_names	List	List of names for each of the N dimensions in psf_flux	required
dimension_units	List	List of astropy.unit.Quantity's describing units of each dimension	required
pixel_size	Quantity	True detector pixel size in dimensions of length/pixel	required
sub_pixel_size	Quantity	PSF file pixel size in dimensions of length/pixel	required
transpose	bool	Whether to transpose the input data in the column/row axis.	False
freeze_dictionary	Dict	Dictionary of dimensions to freeze	{}
check_bounds	bool	Whether to check if the inputs are within the bounds of the PSF model and have the right units. This check causes a small slowdown.	True
extrapolate	bool	Whether to allow the PSF to be evaluated outside of the bounds (i.e. will extrapolate)	False
scale	int	How much to scale the PSF grid. Scale of 2 makes the PSF 2x broader. Default is 1.	1
bin	int	('Optional amount to bin the input PSF file by. Binning the PSF file will result in faster computation, but less accurate modeling. Default is 1.',)	1

trace_pixel property

Pixels corresponding to self.trace_sensitivity

trace_sensitivity property

The sensitivity of a spectrum trace at the self.trace_wavelength and self.trace_pixel positions.

trace_wavelength property

Wavelengths corresponding to self.trace_sensitivity

freeze_dimension(**kwargs)

Drop a dimension of the PSF model. Dropped dimension must exist in self.dimension_names.

Examples:

Freeze a PSF object p to have row of 10 pixels and wavelength of 1.3 microns

>>> p.freeze_dimension(row=10*u.pixel, wavelength=1.3*u.micron)

Returns:

Name	Type	Description
new	PSF	New PSF object

from_file(name, file='/home/runner/work/pandora-psf/pandora-psf/src/pandorapsf/data/pandora_vis_2024-05.fits', transpose=False, extrapolate=False, scale=1, bin=1) classmethod

Build a PSF class from an input fits file.

Parameters:

file: str or fits.HDUList Filename of PSF fits file, or opened hdulist. PSF cube must have a shape such that the first two dimensions are the x and y extent of the PSF. Defaults to visible PSF. transpose: bool Transpose the input file, i.e. rotate 90 degrees

Returns:

psf : pandorapsf.PSF Returns a PSF object

from_name(name, transpose=False, scale=1, bin=1) staticmethod

Open a PSF file based on the detector name. This will automatically freeze dimensions.

Parameters:

Name	Type	Description	Default
name	str	Name of detector	required
transpose	bool	Whether to transpose the input data in the column/row axis.	False
scale	int	How much to scale the PSF grid. Scale of 2 makes the PSF 2x broader. Default is 1.	1
bin	int	('Optional amount to bin the input PSF file by. Binning the PSF file will result in faster computation, but less accurate modeling. Default is 1.',)	1

integrate_spectrum(wavelength, spectrum, wavelength_grid=None)

Create an integrated spectrum, integrated over wavelength bounds.

Parameters:

Name	Type	Description	Default
wavelength	ndarray	Array of wavelength values. If no wavelength unit is present, assumed to be in microns.	required
spectrum	ndarray	Array of spectrum intensity values. Should have units of erg / (Angstrom s cm2).	required
wavelength_grid	ndarray	Optional wavelength grid at which to integrate. If no grid is provided, will default to the trace_wavelength grid of this object.	None

integrate_wavelength(teff=5500, logg=4.5, **kwargs)

Inte

Parameters:

Name	Type	Description	Default
teff		Effective temperature of a star in K.	5500
logg		Surface gravity of a star. log(g)	4.5

prf(row, column, gradients=False, **kwargs)

Bins the PSF down to the pixel scale.

Returns:

Name	Type	Description
row_array	NDArray	Array of integer row positions
column_array	NDArray	Array of integer column positions
prf	ndarray	2D array of PRF flux values with shape (nrows, ncolumns)

Parameters:

Name	Type	Description	Default
row		Row position on the detector	required
column		Column position on the detector	required
gradients		Whether to return gradients. If True, will return an additional 2 arrays that contain the gradients in each axis.	False

psf(gradients=False, **kwargs)

Interpolate the PSF cube to a particular point

Returns:

Name	Type	Description
ar	np.ndarray of shape self.shape	The interpolated PSF

Parameters:

Name	Type	Description	Default
gradients		Whether to return gradients. If True, will return an additional 2 arrays that contain the gradients in each axis.	False