Re: Treatment of uncertainties from the TFractionFitter

Hi Ashish,

|TFraction| fitter will only give you statistical error. If, e.g., your MC is less than perfect model for the shapes in data that error will not/cannot be included by fitter. If you look at some goodness-of-fit (like chisq) systematics from wrong shapes may show up as a bad fit.

To get systematic error you need to do something external to fit. E.g., vary the shapes and see how much results change. |TFractionFitter| does provide a way to apply weights to each histogram your fitting to so you can vary there shapes.

If your shape variations make goodness-of-fit get too bad, you can plausible argue that that variation was too big and use something smaller.

Beyond that there's nothing much in the way of a systematics way to estimate systematic uncertainties; the variations people use tend to be pretty arbitary ...

-Art S.

A.E. Snyder, Group EC                        \!c*p?/
SLAC Mail Stop #95                          ((.   .))
Box 4349                                        |
Stanford, Ca, USA, 94309                      '\|/`
e-mail:snyder_at_slac.stanford.edu                 o
phone:650-926-2701                              _
http://www.slac.stanford.edu/~snyder          BaBar
FAX:650-926-2657                          Collaboration

On Thu, 1 Apr 2010, Ashish Kumar wrote:

> Hello,
> I am using TFractionFitter to fit the data distribution with the
> templates from Monte Carlo simulation to extract their respective
> fractions. The Fitter gives us the fractions with the errors. The
> problem is how to disentangle the error into statistics and systematics
> parts. The error must have some part as statistical, but, clearly it has
> systematics too since it varies the shape of the templates while
> fitting. Can you please help me with this?
>
> Thanks.
> Ashish
>
>
Received on Thu Apr 01 2010 - 19:30:35 CEST