onlinemeas fits have convergence issues

[kostrzewa]

The PP and AP fits implemented in onlinemeas do not converge properly in some (many?) cases.

As an example fits on cA211.15.48 and cA211.15.64 are shown here:

The fit results (blue and red lines and error rectangles) are just barely visible above the effective mass of the PP correlator from the online measurement data.

Using the same data, the same boot.l and the same boot.R but fitting with matrixfit instead results in very reasonable fits:

The chi^2 in the matrixfit on cA211.15.64 is on the high side.