We want to combine these m j measurements to derive n

Info icon This preview shows pages 54–58. Sign up to view the full content.

We want to combine these ( M × J ) measurements to derive N coefficients. We denote the matrix of measured parameters by X , which is ( M × J ) [using conventional matrix notation, not IDL notation, in which in the vertical dimension (number of rows) is M and the horizontal dimension (number of columns) is J ]. We will want to concatenate this matrix to a one-dimensional vector x of length ( MJ ). Specifically, the M × J datapoint matrix is
Image of page 54

Info icon This preview has intentionally blurred sections. Sign up to view the full version.

– 55 – Conventional Fit -3 -2 -1 0 1 2 3 t -3 -2 -1 0 1 2 3 4 y Proper Fit -3 -2 -1 0 1 2 3 t -3 -2 -1 0 1 2 3 4 y Fig. 14.1.— Comparing a conventional fit for to y = a 0 + a 1 t (left panel) to a proper one when both measured variables have errors. On the right, the ellipses denote the correlated errors in ( t, y ); these are the generalization of the errorbars on the left. The right-hand slope is a bit steeper than the left-hand one. X = x 0 , 0 x 0 , 1 x 0 , 2 . . . x 0 ,J - 1 x 1 , 0 x 1 , 1 x 1 , 2 . . . x 1 ,J - 1 x 2 , 0 x 2 , 1 x 2 , 2 . . . x 2 ,J - 1 . . . . . . . . . . . . . . . . . . . . . x M - 1 , 0 x M - 1 , 1 x M - 1 , 2 . . . x M - 1 ,J - 1 (14.4a) We don’t use this big matrix in the solution. Instead, we turn it into a vector in which the first J elements are the data for m = 0, the second J elements are for m = 1, etc. So the vector has dimensions ( MJ ) × 1, like this:
Image of page 55
– 56 – x = x 0 , 0 x 0 , 1 x 0 , 2 . . . x 0 ,J - 1 x 1 , 0 x 1 , 1 x 1 , 2 . . . x 1 ,J - 1 x 2 , 0 x 2 , 1 x 2 , 2 . . . x 2 ,J - 1 . . . . . . . . . x M - 1 , 0 x M - 1 , 1 x M - 1 , 2 . . . x M - 1 ,J - 1 (14.4b) One important reason for writing the whole set of data as a vector instead of a matrix is to make it possible to write the covariance matrix for all measured data in the conventional form, as we now discuss.
Image of page 56

Info icon This preview has intentionally blurred sections. Sign up to view the full version.

– 57 – 14.4. The Data Covariance Matrix and Defining Chi-Square The whole object of fitting is to minimize the chi-square. When all measured parameters have uncertainties, their uncertainties can be correlated. We have to generalize the definition of chi-square accordingly. First, suppose that the observational errors in the datapoints are uncorrelated. Then the intrinsic variance of each datapoint is described by a single number. In our example, for uncorrelated observational errors we’d have the variances in the y values be ( σ 2 y 0 , σ 2 y 1 , . . . ), and similarly for the t values; this would give χ 2 = summationdisplay m δy 2 m σ 2 ym + δt 2 m σ 2 tm (14.5) However, it is common that errors are correlated. For example, if we were fitting y to a polynomial in t , then the errors in the various powers of t would certainly be correlated. More generally, then, the covariances among the different measured values are nonzero. These covariances are the off-diagonal terms in the covariance matrix. Thus, if we denote the covariance matrix for
Image of page 57
Image of page 58
This is the end of the preview. Sign up to access the rest of the document.

{[ snackBarMessage ]}

What students are saying

  • Left Quote Icon

    As a current student on this bumpy collegiate pathway, I stumbled upon Course Hero, where I can find study resources for nearly all my courses, get online help from tutors 24/7, and even share my old projects, papers, and lecture notes with other students.

    Student Picture

    Kiran Temple University Fox School of Business ‘17, Course Hero Intern

  • Left Quote Icon

    I cannot even describe how much Course Hero helped me this summer. It’s truly become something I can always rely on and help me. In the end, I was not only able to survive summer classes, but I was able to thrive thanks to Course Hero.

    Student Picture

    Dana University of Pennsylvania ‘17, Course Hero Intern

  • Left Quote Icon

    The ability to access any university’s resources through Course Hero proved invaluable in my case. I was behind on Tulane coursework and actually used UCLA’s materials to help me move forward and get everything together on time.

    Student Picture

    Jill Tulane University ‘16, Course Hero Intern