vis.speed.fit_nospeed

  FIT_NOSPEED - Fit a 2D Gaussian with no speed parameter (xi=0).

   [PARAMETERS, SSE, R_SQUARED] = vis.speed.fit_nospeed(SF, TF, R, ...)

   Fits a Gaussian to a set of responses where the speed parameter 'xi'
   is constrained to be 0. This function uses a harmonized multi-start
   approach that is identical to `vis.speed.fit` to ensure a fair comparison
   for nested F-tests.

   Inputs:
     sf - An array of spatial frequency values.
     tf - An array of temporal frequency values.
     r  - An array of measured responses.

   Name-Value Pair Arguments:
      numberStartPoints  - The number of initial starting points to try. (Default: 40)
      SpecificStartPoint - A 7xN matrix specifying specific starting
                           points to include in the search.

   Outputs:
     PARAMETERS - A 7x1 vector with the best-fit parameters.
     SSE        - The total sum of squared errors for the best fit.
     R_SQUARED  - The R-squared value (coefficient of determination).

   Parameters:
     [A, zeta, xi, sigma_sf, sigma_tf, sf0, tf0], where xi is always 0.

   See also: vis.speed.fit, lsqcurvefit