pyimcom.meta.distortimage

Metadetection mosaic class and associated functions.

Classes

MetaMosaic: A (sub-)mosaic that can be sheared.

Functions

shearimage_to_fits: Writes a sheared image to a FITS file.

Classes

MetaMosaic

Contains mosaic information for use in meta operations.

Functions

shearimage_to_fits(im, fname[, layers, overwrite])

Utility to save a shearimage dictionary a FITS file.

Module Contents

class MetaMosaic(fname, bbox=None, extpix=None, verbose=False)[source]

Contains mosaic information for use in meta operations.

Parameters:

fname (str) – File name of block to build (sub-mosaic will be a 3x3 block region centered on this).
bbox (list of int, optional) – Boundaries [xmin, xmax, ymin, ymax]: accept only blocks with xmin<=x<xmax, ymin<=y<ymax. Default is to read the whole mosaic. This feature is useful if you haven’t run the whole mosaic or pulled it to your disk.
extpix (int, optional) – Number of pixels to extend beyond the block boundary. Uses full 3x3 submosaic if not given.
verbose (bool, optional) – Whether to print diagnostics.

cfg[source]

PyIMCOM configuration used in this file.

Type:: pyimcom.config.Config

Nside[source]

Side length of moasic image.

Type:: int

in_image[source]

3D array; the input image cube.

Type:: np.array

in_fidelity[source]

2D fidelity map

Type:: np.array of float

in_noise[source]

2D noise map

Type:: np.array of float

in_mask[source]

2D mask

Type:: np.array of bool

wcs[source]

The WCS for the mosaic.

Type:: astropy.wcs.WCS

__init__()[source]: Constructor.

maskpix()[source]: Masks an additional set of pixels.

mask_fidelity_cut()[source]: Implements a mask based on PyIMCOM fidelity.

mask_noise_cut()[source]: Implements a mask based on PyIMCOM noise.

mask_caps()[source]: Implements a mask based on a catalog of circles.

in_mask()[source]: Boolean mask for input data (False = OK, True = masked).

to_file()[source]: Writes the mosaic object to a FITS file.

shearimage()[source]: Generate a sheared image.

origimage()[source]: Same format as shearimage, but doesn’t do shear/reconvolution (just extracts subarrays).

cfg[source]

LegacyName = False[source]

cprfitsgz = False[source]

stem[source]

ix[source]

iy[source]

Nside[source]

in_image[source]

in_fidelity[source]

in_noise[source]

in_mask[source]

wcs[source]

maskpix(extramask)[source]

Provide additional masking beyond the default.

Pixels that are True in extramask are masked out.

Parameters:: extramask (np.array of bool) – 2D array of pixels to mask out.
Return type:: None

mask_fidelity_cut(fidelitymin)[source]

Masks pixels with fidelity below the cut.

For example, mosaic.mask_fidelity_cut(40) will cut pixels with leakage worse than 40 dB (i.e. U/C>1e-4).

Parameters:: fidelitymin (float) – The fidelity cut in dB.
Return type:: None

mask_noise_cut(noisemax)[source]

Masks pixels with noise metric worse than the cut.

For example, mosaic.mask_noise_cut(3) will cut pixels whose noise suppression is less than 3 dB relative to input (i.e., Sigma>10**(-0.3)).

Parameters:: noisemax (float) – The noise cut in dB.
Return type:: None

mask_caps(ra, dec, radius)[source]

Masks circular regions around an input catalog.

Each of ra, dec, and radius should be an array-like object of the same length N (radius is permitted to be a scalar).

Parameters:

ra (np.array of float) – The right ascensions of the objects to mask (in degrees).
dec (np.array of float) – The declinations of the objects to mask (in degrees).
radius (float or np.array of float) – The radius to mask around each object (in degrees). If a scalar value is given, uses the same radius for all objects.

Return type:

None

to_file(fname)[source]

Writes the input mosaic images to a FITS file.

Parameters:: fname (str) – File name for the output FITS file.
Return type:: None

shearimage(N, jac=None, psfgrow=1.0, oversamp=1.0, fidelity_min=None, Rsearch=6.0, select_layers=None, verbose=False)[source]

Generates a sheared image and its WCS.

Parameters:

N (int) – Size of the output image (shape will be (N, N)).
jac (np.array or None, optional) – 2x2 Jacobian for transformation (None defaults to the identity).
psfgrow (float, optional) – Factor (in linear scale) by which to grow the PSF.
oversamp (float, optional) – Up-sampling factor (e.g., 1 = preserve pixel scale).
fidelity_min (float or None, optional) – Fidelity cut (in dB) for which pixels to use.
Rsearch (float, optional) – Search radius in interpolation, in units of coadded pixels.
select_layers (np.array of int or None, optional) – If given, only process these layers.
stest (int, optional) – Computes test diagnostics (leakage & noise for interpolation) every this many output pixels. Set to 1 to do every output pixel (but this may be slower).
verbose (bool, optional) – Print diagnostics to terminal.

Returns:

im – Image dictionary containing ‘image’, ‘mask’, ‘wcs’, ‘pars’, ‘layers’, ‘psf_fwhm’, and ‘ref’.

Return type:

dict

Notes

The output image contains: * im['image'] : np.array, image cube (3D) * im['mask'] : np.array, image mask (2D, Boolean, True=masked) * im['wcs'] : astropy.wcs.WCS, world coordinate system object (appropriate for a FITS file) * im['pars'] : dict, parameter dictionary (can be turned into a FITS header) * im['layers'] : list of str, names of layers * im['psf_fwhm'] : float, 1 sigma width of output PSF in arcsec * im['ref'] : (int, int), projection center (x,y), 0-offset convention

The sense of jac is that the output is related to the input by: d{input coords[i]} = sum_j jac[i,j] d{output coords[j]}

Assumes a Gaussian PSF, returns an error if something else is used.

origimage(N=None, select_layers=None)[source]

Like shearimage, but without applying the deconvolution/shear/reconvolution (so is faster).

Parameters:

N (int, optional) – Size of the output image (shape will be (N, N)). Default is one block.
select_layers (np.array of int or None, optional) – If given, only process these layers.

Returns:

im – Image dictionary

Return type:

dict