Open flipdazed opened 8 years ago
It is assumed the function gW
has no error but the success of gW
is coupled to the number of autocorrelations.
A start would be to plot the errors in gW
and then determine the uncertainty in the first intersection of the x-axis
Issue The
uWerr
routine is great but in some cases the windowing function picks up very bad estimates for the integration window. This is not reflected in the error bars and consequently presents misleading figures.Resolution Identify a method of robustly including the uncertainty in the windowing procedure into the overall error bar estimation