6.2.7.4. UsineFit@Outputs

USINE outputs enabled after minimization runs.

Note

The best-fit parameters are always stored in the file fit_result.out (in the user-chosen directory for the run).

Group@subgroup@parameter Description
UsineFit@Outputs@IsPrintCovMatrix (M=0) If true, save the covariance matrix of the best-fit parameters (FIT and NUISANCE) obtained by the minimiser in fit_covmatrix.out (in the user-chosen directory for the run).
UsineFit@Outputs@IsPrintHessianMatrix (M=0) Same, but for Hessian matrix (inverse of covariance matrix), saved in fit_hessianmatrix.out.
UsineFit@Outputs@Contours (M=1) Syntax in initialisation file is X1,X2:N:{CL1,CL2...}, with X1 and X2 the parameter names (use keyword ALL to get all contour combinations of fit parameters), N the number of points to use in the contour, CLi the contours to draw in unit of \(\sigma\). The output is a plot of X1 vs X2, also saved in the file fit_contours_X1_vs_X2.out (in the user-chosen directory for the run). Note that extracting contours can be very slow and sometimes it fails to get (and thus plot) the sought contours. Only use at the last steps of your analysis, and for well-behaved minima (i.e., \(\chi^2_{\rm min}/{\rm dof}\sim 1\)).
UsineFit@Outputs@Profile (M=1) Syntax in initialisation file is X:N, with X the parameter name (use keyword ALL to scan all fit parameters), and N the number of points to use in the scan. The output is a plot of \(\chi_{\rm min}^2(X)\), also saved in the file fit_profilelikelihood_X.out (in the user-chosen directory for the run). This can take some time (do not confuse it with the Scan keyword above) for the following reason: imagine that your initial fit had \(n_{\rm fit}\) parameters, now you need to find the new minimum (for each parameter) at N positions, minimising on the \(n_{\rm fit}-1\) other parameters. Note that if your initial minimisation (\(n_{\rm fit}\) parameters) converged well, this is guaranteed to work well as well.
UsineFit@Outputs@Scan (M=1) Syntax in initialisation file is X:N, with X the parameter name (use keyword ALL to scan all fit parameters), and N the number of points to use in the scan. The output is a plot of \(\chi^2(X)\), also saved in the file fit_parameterscan_X.out (in the user-chosen directory for the run). This takes no time at all, as there is no further minimisation (do not confuse it with the Profile keyword above), only N calls to evaluate \(\chi^2\) at various positions of the parameter.