computechi2¶

class pydl.pydlutils.math.computechi2(bvec, sqivar, amatrix)[source]¶

Bases: object

Solve the linear set of equations \(A x = b\) using SVD.

The attributes of this class are all read-only properties, implemented with lazyproperty.

Parameters

bvecnumpy.ndarray: The \(b\) vector in \(A x = b\). This vector has length \(N\).
sqivarnumpy.ndarray: The reciprocal of the errors in bvec. The name comes from the square root of the inverse variance, which is what this is.
amatrixnumpy.ndarray: The matrix \(A\) in \(A x = b\). The shape of this matrix is (\(N\), \(M\)).

Initialize the object and perform initial computations.

Attributes Summary

`acoeff`	(`ndarray`) The fit parameters, \(x\), in \(A x = b\).
`chi2`	(`float`) The \(\chi^2\) value of the fit.
`covar`	(`ndarray`) The covariance matrix.
`dof`	(`int`) The degrees of freedom of the fit.
`var`	(`ndarray`) The variances of the fit.
`yfit`	(`ndarray`) The evaluated best-fit at each point.

Attributes Documentation

acoeff¶: (ndarray) The fit parameters, \(x\), in \(A x = b\). This vector has length \(M\).

covar¶: (ndarray) The covariance matrix. The shape of this matrix is (\(M\), \(M\)).

dof¶: (int) The degrees of freedom of the fit. This is the number of values of bvec that have sqivar > 0 minus the number of fit paramaters, which is equal to \(M\).

var¶: (ndarray) The variances of the fit. This is identical to the diagonal of the covariance matrix. This vector has length \(M\).

yfit¶: (ndarray) The evaluated best-fit at each point. This vector has length \(N\).