PsfMatchConfigDF#

class lsst.ip.diffim.PsfMatchConfigDF(*args, **kw)#

Bases: PsfMatchConfig

The parameters specific to the delta-function (one basis per-pixel) Psf-matching basis

Attributes Summary

`afwBackgroundConfig`	Controlling the Afw background fitting (`SubtractBackgroundConfig`, default `<class 'lsst.meas.algorithms.subtractBackground.SubtractBackgroundConfig'>`)
`badMaskPlanes`	Mask planes to ignore when calculating diffim statistics Options: NO_DATA EDGE SAT BAD CR INTRP (`List`, default `('NO_DATA', 'EDGE', 'SAT')`)
`calculateKernelUncertainty`	Calculate kernel and background uncertainties for each kernel candidate? This comes from the inverse of the covariance matrix.
`candidateCoreRadius`	Radius for calculation of stats in 'core' of KernelCandidate diffim.
`candidateResidualMeanMax`	Rejects KernelCandidates yielding bad difference image quality.
`candidateResidualStdMax`	Rejects KernelCandidates yielding bad difference image quality.
`centralRegularizationStencil`	Type of stencil to approximate central derivative (for centralDifference only) (`int`, default `9`)
`checkConditionNumber`	Test for maximum condition number when inverting a kernel matrix.
`conditionNumberType`	Use singular values (SVD) or eigen values (EIGENVALUE) to determine condition number (`str`, default `'EIGENVALUE'`)
`constantVarianceWeighting`	Use constant variance weighting in single kernel fitting? In some cases this is better for bright star residuals.
`fitForBackground`	Include terms (including kernel cross terms) for background in ip_diffim (`bool`, default `False`)
`forwardRegularizationOrders`	Array showing which order derivatives to penalize (for forwardDifference only) (`List`, default `(1, 2)`)
`iterateSingleKernel`	Remake KernelCandidate using better variance estimate after first pass? Primarily useful when convolving a single-depth image, otherwise not necessary.
`kernelBasisSet`	Type of basis set for PSF matching kernel.
`kernelSize`	Number of rows/columns in the convolution kernel; should be odd-valued.
`kernelSizeFwhmScaling`	Multiplier of the largest AL Gaussian basis sigma to get the kernel bbox (pixel) size.
`kernelSizeMax`	Maximum kernel bbox (pixel) size.
`kernelSizeMin`	Minimum kernel bbox (pixel) size.
`kernelSumClipping`	Do sigma clipping on the ensemble of kernel sums (`bool`, default `True`)
`lambdaMax`	If scan through lambda needed (minimizeBiasedRisk, minimizeUnbiasedRisk), stop at this value.
`lambdaMin`	If scan through lambda needed (minimizeBiasedRisk, minimizeUnbiasedRisk), start at this value.
`lambdaScaling`	Fraction of the default lambda strength (N.R.
`lambdaStep`	If scan through lambda needed (minimizeBiasedRisk, minimizeUnbiasedRisk), step in these increments.
`lambdaStepType`	If a scan through lambda is needed (minimizeBiasedRisk, minimizeUnbiasedRisk),
`lambdaType`	How to choose the value of the regularization strength (`str`, default `'absolute'`)
`lambdaValue`	Value used for absolute determinations of regularization strength (`float`, default `0.2`)
`maxConditionNumber`	Maximum condition number for a well conditioned matrix (`float`, default `50000000.0`)
`maxKsumSigma`	Maximum allowed sigma for outliers from kernel sum distribution.
`maxSpatialConditionNumber`	Maximum condition number for a well conditioned spatial matrix (`float`, default `10000000000.0`)
`maxSpatialIterations`	Maximum number of iterations for rejecting bad KernelCandidates in spatial fitting (`int`, default `3`)
`nStarPerCell`	Maximum number of KernelCandidates in each SpatialCell to use in the spatial fitting.
`numPrincipalComponents`	Number of principal components to use for Pca basis, including the mean kernel if requested.
`regularizationBorderPenalty`	Value of the penalty for kernel border pixels (`float`, default `3.0`)
`regularizationType`	Type of regularization.
`scaleByFwhm`	Scale kernelSize, alardGaussians by input Fwhm (`bool`, default `True`)
`singleKernelClipping`	Do sigma clipping on each raw kernel candidate (`bool`, default `True`)
`sizeCellX`	Size (rows) in pixels of each SpatialCell for spatial modeling (`int`, default `128`)
`sizeCellY`	Size (columns) in pixels of each SpatialCell for spatial modeling (`int`, default `128`)
`spatialBgOrder`	Spatial order of differential background variation (`int`, default `1`)
`spatialKernelClipping`	Do sigma clipping after building the spatial model (`bool`, default `True`)
`spatialKernelOrder`	Spatial order of convolution kernel variation (`int`, default `2`)
`spatialModelType`	Type of spatial functions for kernel and background (`str`, default `'chebyshev1'`)
`subtractMeanForPca`	Subtract off the mean feature before doing the Pca (`bool`, default `True`)
`useAfwBackground`	Use afw background subtraction instead of ip_diffim (`bool`, default `False`)
`useBicForKernelBasis`	Use Bayesian Information Criterion to select the number of bases going into the kernel (`bool`, default `False`)
`useCoreStats`	Use the core of the footprint for the quality statistics, instead of the entire footprint.
`usePcaForSpatialKernel`	Use Pca to reduce the dimensionality of the kernel basis sets.
`useRegularization`	Use regularization to smooth the delta function kernels (`bool`, default `True`)
`warpingConfig`	Config for warping exposures to a common alignment (`WarperConfig`, default `<class 'lsst.afw.math._warper.WarperConfig'>`)

Methods Summary

setDefaults()

Subclass hook for computing defaults.

Attributes Documentation

afwBackgroundConfig#: Controlling the Afw background fitting (SubtractBackgroundConfig, default <class 'lsst.meas.algorithms.subtractBackground.SubtractBackgroundConfig'>)

badMaskPlanes#: Mask planes to ignore when calculating diffim statistics Options: NO_DATA EDGE SAT BAD CR INTRP (List, default ('NO_DATA', 'EDGE', 'SAT'))

calculateKernelUncertainty#: Calculate kernel and background uncertainties for each kernel candidate? This comes from the inverse of the covariance matrix. Warning: regularization can cause problems for this step. (bool, default False)

candidateCoreRadius#: Radius for calculation of stats in ‘core’ of KernelCandidate diffim. Total number of pixels used will be (2*radius)**2. This is used both for ‘core’ diffim quality as well as ranking of KernelCandidates by their total flux in this core (int, default 3)

candidateResidualMeanMax#: Rejects KernelCandidates yielding bad difference image quality. Used by BuildSingleKernelVisitor, AssessSpatialKernelVisitor. Represents average over pixels of (image/sqrt(variance)). (float, default 0.25)

candidateResidualStdMax#: Rejects KernelCandidates yielding bad difference image quality. Used by BuildSingleKernelVisitor, AssessSpatialKernelVisitor. Represents stddev over pixels of (image/sqrt(variance)). (float, default 1.5)

centralRegularizationStencil#

Type of stencil to approximate central derivative (for centralDifference only) (int, default 9)

Allowed values:

'5': 5-point stencil including only adjacent-in-x,y elements
'9': 9-point stencil including diagonal elements
'None': Field is optional

checkConditionNumber#: Test for maximum condition number when inverting a kernel matrix. Anything above maxConditionNumber is not used and the candidate is set as BAD. Also used to truncate inverse matrix in estimateBiasedRisk. However, if you are doing any deconvolution you will want to turn this off, or use a large maxConditionNumber (bool, default False)

conditionNumberType#

Use singular values (SVD) or eigen values (EIGENVALUE) to determine condition number (str, default 'EIGENVALUE')

Allowed values:

'SVD': Use singular values
'EIGENVALUE': Use eigen values (faster)
'None': Field is optional

constantVarianceWeighting#: Use constant variance weighting in single kernel fitting? In some cases this is better for bright star residuals. (bool, default True)

fitForBackground#: Include terms (including kernel cross terms) for background in ip_diffim (bool, default False)

forwardRegularizationOrders#: Array showing which order derivatives to penalize (for forwardDifference only) (List, default (1, 2))

iterateSingleKernel#: Remake KernelCandidate using better variance estimate after first pass? Primarily useful when convolving a single-depth image, otherwise not necessary. (bool, default False)

kernelBasisSet#

Type of basis set for PSF matching kernel. (str, default 'alard-lupton')

Allowed values:

'alard-lupton'

Alard-Lupton sum-of-gaussians basis set,

The first term has no spatial variation
The kernel sum is conserved
You may want to turn off ‘usePcaForSpatialKernel’

'delta-function'

Delta-function kernel basis set,

You may enable the option useRegularization
You should seriously consider usePcaForSpatialKernel, which will also enable kernel sum conservation for the delta function kernels

'None'

Field is optional

kernelSize#: Number of rows/columns in the convolution kernel; should be odd-valued. Modified by kernelSizeFwhmScaling if scaleByFwhm = true (int, default 21)

kernelSizeFwhmScaling#: Multiplier of the largest AL Gaussian basis sigma to get the kernel bbox (pixel) size. (float, default 6.0)

kernelSizeMax#: Maximum kernel bbox (pixel) size. (int, default 35)

kernelSizeMin#: Minimum kernel bbox (pixel) size. (int, default 21)

kernelSumClipping#: Do sigma clipping on the ensemble of kernel sums (bool, default True)

lambdaMax#: If scan through lambda needed (minimizeBiasedRisk, minimizeUnbiasedRisk), stop at this value. If lambdaStepType = log:linear, suggest 2:100 (float, default 2.0)

lambdaMin#: If scan through lambda needed (minimizeBiasedRisk, minimizeUnbiasedRisk), start at this value. If lambdaStepType = log:linear, suggest -1:0.1 (float, default -1.0)

lambdaScaling#: Fraction of the default lambda strength (N.R. 18.5.8) to use. 1e-4 or 1e-5 (float, default 0.0001)

lambdaStep#: If scan through lambda needed (minimizeBiasedRisk, minimizeUnbiasedRisk), step in these increments. If lambdaStepType = log:linear, suggest 0.1:0.1 (float, default 0.1)

lambdaStepType#

If a scan through lambda is needed (minimizeBiasedRisk, minimizeUnbiasedRisk),: use log or linear steps (str, default 'log')

Allowed values:

'log': Step in log intervals; e.g. lambdaMin, lambdaMax, lambdaStep = -1.0, 2.0, 0.1
'linear': Step in linear intervals; e.g. lambdaMin, lambdaMax, lambdaStep = 0.1, 100, 0.1
'None': Field is optional

lambdaType#

How to choose the value of the regularization strength (str, default 'absolute')

Allowed values:

'absolute': Use lambdaValue as the value of regularization strength
'relative': Use lambdaValue as fraction of the default regularization strength (N.R. 18.5.8)
'minimizeBiasedRisk': Minimize biased risk estimate
'minimizeUnbiasedRisk': Minimize unbiased risk estimate
'None': Field is optional

lambdaValue#: Value used for absolute determinations of regularization strength (float, default 0.2)

maxConditionNumber#: Maximum condition number for a well conditioned matrix (float, default 50000000.0)

maxKsumSigma#: Maximum allowed sigma for outliers from kernel sum distribution. Used to reject variable objects from the kernel model (float, default 3.0)

maxSpatialConditionNumber#: Maximum condition number for a well conditioned spatial matrix (float, default 10000000000.0)

maxSpatialIterations#: Maximum number of iterations for rejecting bad KernelCandidates in spatial fitting (int, default 3)

nStarPerCell#: Maximum number of KernelCandidates in each SpatialCell to use in the spatial fitting. Set to -1 to use all candidates in each cell. (int, default 5)

numPrincipalComponents#: Number of principal components to use for Pca basis, including the mean kernel if requested. (int, default 5)

regularizationBorderPenalty#: Value of the penalty for kernel border pixels (float, default 3.0)

regularizationType#

Type of regularization. (str, default 'centralDifference')

Allowed values:

'centralDifference': Penalize second derivative using 2-D stencil of central finite difference
'forwardDifference': Penalize first, second, third derivatives using forward finite differeces
'None': Field is optional

scaleByFwhm#: Scale kernelSize, alardGaussians by input Fwhm (bool, default True)

singleKernelClipping#: Do sigma clipping on each raw kernel candidate (bool, default True)

sizeCellX#: Size (rows) in pixels of each SpatialCell for spatial modeling (int, default 128)

sizeCellY#: Size (columns) in pixels of each SpatialCell for spatial modeling (int, default 128)

spatialBgOrder#: Spatial order of differential background variation (int, default 1)

spatialKernelClipping#: Do sigma clipping after building the spatial model (bool, default True)

spatialKernelOrder#: Spatial order of convolution kernel variation (int, default 2)

spatialModelType#

Type of spatial functions for kernel and background (str, default 'chebyshev1')

Allowed values:

'chebyshev1': Chebyshev polynomial of the first kind
'polynomial': Standard x,y polynomial
'None': Field is optional

subtractMeanForPca#: Subtract off the mean feature before doing the Pca (bool, default True)

useAfwBackground#: Use afw background subtraction instead of ip_diffim (bool, default False)

useBicForKernelBasis#: Use Bayesian Information Criterion to select the number of bases going into the kernel (bool, default False)

useCoreStats#: Use the core of the footprint for the quality statistics, instead of the entire footprint. WARNING: if there is deconvolution we probably will need to turn this off (bool, default False)

usePcaForSpatialKernel#: Use Pca to reduce the dimensionality of the kernel basis sets. This is particularly useful for delta-function kernels. Functionally, after all Cells have their raw kernels determined, we run a Pca on these Kernels, re-fit the Cells using the eigenKernels and then fit those for spatial variation using the same technique as for Alard-Lupton kernels. If this option is used, the first term will have no spatial variation and the kernel sum will be conserved. (bool, default False)

useRegularization#: Use regularization to smooth the delta function kernels (bool, default True)

warpingConfig#: Config for warping exposures to a common alignment (WarperConfig, default <class 'lsst.afw.math._warper.WarperConfig'>)

Methods Documentation

setDefaults()#: Subclass hook for computing defaults.

Notes#

Derived Config classes that must compute defaults rather than using the Field instances’s defaults should do so here. To correctly use inherited defaults, implementations of setDefaults must call their base class’s setDefaults.

PsfMatchConfigDF#

Notes#