Hi Adrian,
Adrian T Sindile wrote:
>Hi, Chris!
>
>
>
>>Adrian, I remember you saying you used ROOT for error
>>propagation, so you may want to double-check your code too.
>>
>>
>
>I checked my code and I can say I did NOT use the same histogram twice, so
>my code should be error-free from this point of view.
>The way I did it was to define two other histograms leftMinus and
>leftPlus, for example, avoiding using the same leftThMinus or leftThPlus
>twice:
>
>leftMinus->Add(leftThPlus, leftThMinus, 1./qPlus, -1./qMinus);
>leftPlus->Add(leftThPlus, leftThMinus, 1./qPlus, 1./qMinus);
>leftAsym->Divide(leftMinus, leftPlus);
>
Try jason's GetAsymmetry() function, and I'm pretty sure you will get a
different result. The naming doesn't matter; just that leftThPlus is
used both in the numerator and denominator.
--chris
>
>Then, I propagate errors by hand for the super-ratio and GE/GM.
>So, if my code gave different error bars from the common BLAST codes, this
>is one error my code did NOT have.
>
>Adrian
>
>
This archive was generated by hypermail 2.1.2 : Mon Feb 24 2014 - 14:07:30 EST