lsst.meas.modelfit.priors.priorsContinued Namespace Reference

Classes
class	SemiEmpiricalPrior
class	SoftenedLinearPrior

Functions
def	fitMixture (data, nComponents, minFactor=0.25, maxFactor=4.0, nIterations=20, df=float("inf"))

Variables
	SemiEmpiricalPriorConfig = makeConfigClass(SemiEmpiricalPriorControl)
	SoftenedLinearPriorConfig = makeConfigClass(SoftenedLinearPriorControl)

Function Documentation

fitMixture()

def lsst.meas.modelfit.priors.priorsContinued.fitMixture	(		data,
			nComponents,
			minFactor = 0.25,
			maxFactor = 4.0,
			nIterations = 20,
			df = float("inf")
	)

Fit a ``Mixture`` distribution to a set of (e1, e2, r) data points,
returing a ``MixturePrior`` object.

Parameters
----------
data : numpy.ndarray
    array of data points to fit; shape=(N,3)
nComponents : int
    number of components in the mixture distribution
minFactor : float
    ellipticity variance of the smallest component in the initial mixture,
    relative to the measured variance
maxFactor : float
    ellipticity variance of the largest component in the initial mixture,
    relative to the measured variance
nIterations : int
    number of expectation-maximization update iterations
df : float
    number of degrees of freedom for component Student's T distributions
    (inf=Gaussian).

Definition at line 55 of file priorsContinued.py.

               nIterations=20, df=float("inf")):
    """Fit a ``Mixture`` distribution to a set of (e1, e2, r) data points,
    returing a ``MixturePrior`` object.
 
    Parameters
    ----------
    data : numpy.ndarray
        array of data points to fit; shape=(N,3)
    nComponents : int
        number of components in the mixture distribution
    minFactor : float
        ellipticity variance of the smallest component in the initial mixture,
        relative to the measured variance
    maxFactor : float
        ellipticity variance of the largest component in the initial mixture,
        relative to the measured variance
    nIterations : int
        number of expectation-maximization update iterations
    df : float
        number of degrees of freedom for component Student's T distributions
        (inf=Gaussian).
    """
    components = Mixture.ComponentList()
    rMu = data[:, 2].mean()
    rSigma = data[:, 2].var()
    eSigma = 0.5*(data[:, 0].var() + data[:, 1].var())
    mu = np.array([0.0, 0.0, rMu], dtype=float)
    baseSigma = np.array([[eSigma, 0.0, 0.0],
                          [0.0, eSigma, 0.0],
                          [0.0, 0.0, rSigma]])
    for factor in np.linspace(minFactor, maxFactor, nComponents):
        sigma = baseSigma.copy()
        sigma[:2, :2] *= factor
        components.append(Mixture.Component(1.0, mu, sigma))
    mixture = Mixture(3, components, df)
    restriction = MixturePrior.getUpdateRestriction()
    for i in range(nIterations):
        mixture.updateEM(data, restriction)
    return mixture

Variable Documentation

SemiEmpiricalPriorConfig

lsst.meas.modelfit.priors.priorsContinued.SemiEmpiricalPriorConfig = makeConfigClass(SemiEmpiricalPriorControl)

Definition at line 38 of file priorsContinued.py.

SoftenedLinearPriorConfig

lsst.meas.modelfit.priors.priorsContinued.SoftenedLinearPriorConfig = makeConfigClass(SoftenedLinearPriorControl)

Definition at line 40 of file priorsContinued.py.