next up previous contents
Next: Reliability and sensitivity Up: How COMBINE Works Previous: Limit of harmonic order   Contents


Equivalent sig

Each linear combination is assigned an equivalent sig,

\begin{displaymath}
\mathrm{sig}_{\mathrm{eq}}:=\min\left(\left\vert c_{k}\right\vert^{-\delta_{k}}\mathrm{sig}_{k}\right)-\chi \log K\: ,
\end{displaymath} (39)

where $\delta_{k}$ denotes the decay parameter provided by the keyword decay, and $\chi $ is the combination damping, specified using the keyword cdamp. Both keywords are followed by floating-point numbers. The default values for both parameters are $1$.

Figure 1: Ratio of equivalent sig over sig of an individual signal component vs. polynomial coefficient $c_k$ associated to the signal component. Five graphs for different values of the decay parameter $\delta _k$ are presented.
\includegraphics[clip,angle=0,width=110mm, clip]{eps/COMB_decay.eps}

Fig.1 displays the relative sig correction with increasing coefficient $c_k$ for five different values of the decay parameter $\delta _k$. Fig.2 illustrates the correction of equivalent sig with increasing number of components contributing to a linear combination $K$ for five different values of the combination damping $\chi $.

Figure 2: Additive significance correction for a linear combination employing $K$ different signal components. Five graphs for different values of the combination damping $\chi $ are presented.
\includegraphics[clip,angle=0,width=110mm, clip]{eps/COMB_cdamp.eps}


next up previous contents
Next: Reliability and sensitivity Up: How COMBINE Works Previous: Limit of harmonic order   Contents
Piet Reegen 2009-09-23