lsst.meas.extensions.piff.piffPsfDeterminer Namespace Reference

Classes
class	PiffPsfDeterminerConfig
class	PiffPsfDeterminerTask

Functions
def	computeWeight (maskedImage, maxSNR)
def	applyMaxSNR (imArr, weightArr, good, maxSNR)

Function Documentation

applyMaxSNR()

def lsst.meas.extensions.piff.piffPsfDeterminer.applyMaxSNR	(		imArr,
			weightArr,
			good,
			maxSNR
	)

Rescale weight of bright stars to cap the computed SNR.

Parameters
----------
imArr : `ndarray`
    Signal (image) array of stamp.
weightArr : `ndarray`
    Weight map array.  May be rescaled in place.
good : `ndarray`
    Index array of pixels to use when computing SNR.
maxSNR : `float`
    Threshold for adjusting variance plane implementing maximum SNR.

Definition at line 100 of file piffPsfDeterminer.py.

def applyMaxSNR(imArr, weightArr, good, maxSNR):
    """Rescale weight of bright stars to cap the computed SNR.
 
    Parameters
    ----------
    imArr : `ndarray`
        Signal (image) array of stamp.
    weightArr : `ndarray`
        Weight map array.  May be rescaled in place.
    good : `ndarray`
        Index array of pixels to use when computing SNR.
    maxSNR : `float`
        Threshold for adjusting variance plane implementing maximum SNR.
    """
    # We define the SNR value following Piff.  Here's the comment from that
    # code base explaining the calculation.
    #
    # The S/N value that we use will be the weighted total flux where the
    # weight function is the star's profile itself.  This is the maximum S/N
    # value that any flux measurement can possibly produce, which will be
    # closer to an in-practice S/N than using all the pixels equally.
    #
    # F = Sum_i w_i I_i^2
    # var(F) = Sum_i w_i^2 I_i^2 var(I_i)
    #        = Sum_i w_i I_i^2             <--- Assumes var(I_i) = 1/w_i
    #
    # S/N = F / sqrt(var(F))
    #
    # Note that if the image is pure noise, this will produce a "signal" of
    #
    # F_noise = Sum_i w_i 1/w_i = Npix
    #
    # So for a more accurate estimate of the S/N of the actual star itself, one
    # should subtract off Npix from the measured F.
    #
    # The final formula then is:
    #
    # F = Sum_i w_i I_i^2
    # S/N = (F-Npix) / sqrt(F)
    F = np.sum(weightArr[good]*imArr[good]**2, dtype=float)
    Npix = np.sum(good)
    SNR = 0.0 if F < Npix else (F-Npix)/np.sqrt(F)
    # rescale weight of bright stars.  Essentially makes an error floor.
    if SNR > maxSNR:
        factor = (maxSNR / SNR)**2
        weightArr[good] *= factor
 
 

computeWeight()

def lsst.meas.extensions.piff.piffPsfDeterminer.computeWeight	(		maskedImage,
			maxSNR
	)

Derive a weight map without Poisson variance component due to signal.

Parameters
----------
maskedImage : `afw.image.MaskedImage`
    PSF candidate postage stamp
maxSNR : `float`
    Maximum SNR applying variance floor.

Returns
-------
weightArr : `ndarry`
    Array to use for weight.

Definition at line 69 of file piffPsfDeterminer.py.

def computeWeight(maskedImage, maxSNR):
    """Derive a weight map without Poisson variance component due to signal.
 
    Parameters
    ----------
    maskedImage : `afw.image.MaskedImage`
        PSF candidate postage stamp
    maxSNR : `float`
        Maximum SNR applying variance floor.
 
    Returns
    -------
    weightArr : `ndarry`
        Array to use for weight.
    """
    imArr = maskedImage.image.array
    varArr = maskedImage.variance.array
    good = (varArr != 0) & np.isfinite(varArr) & np.isfinite(imArr)
 
    # Fit a straight line to variance vs (sky-subtracted) signal.
    # The evaluate that line at zero signal to get an estimate of the
    # signal-free variance.
    fit = np.polyfit(imArr[good], varArr[good], deg=1)
    # fit is [1/gain, sky_var]
    weightArr = np.zeros_like(imArr, dtype=float)
    weightArr[good] = 1./fit[1]
 
    applyMaxSNR(imArr, weightArr, good, maxSNR)
    return weightArr