Module to find the norm of a signal and to return a normaliseLSOd array. The original signal is left untouched. More...
Prototypes | |
| void | LALInspiralWaveNormaliseLSO (LALStatus *status, REAL4Vector *filter, REAL8 *norm, InspiralWaveNormaliseIn *in) |
Module to find the norm of a signal and to return a normaliseLSOd array. The original signal is left untouched.
Given the positive frequency Fourier components \(H_k,\) \(k=0,\ldots,n-1,\) of a vector and the noise PSD \(S_m,\) \(m=0,\ldots,n/2,\) this module first computes the norm \(H\) of the vector treating \(S_m\) as the measure: (note that in {fftw} notation, the zeroth frequency component is \(H_0,\) Nyquist is \(H_{n/2},\) \(H_k,\) \(k \ne 0,n/2,\) ( \(H_{n-k})\) is the real (imaginary) part of the \(k\)th harmonic)
\begin{equation} \label{eq_inspiralnorm} H = \sum_{k=1}^{n/2-1} \frac{H_k^2 + H^2_{n-k}}{S_k}. \end{equation}
The above sum is limited to frequency in->fCutoff. Also, note that the zeroth and Nyquist frequency components are ignored in the computation of the norm. Moreover, array elements of filter corresponding to frequencies greater than in->fCutoff are set to zero. That is, the code replaces the original vector \(H_k\) with normalized vector using:
\begin{eqnarray} \widehat H_k & = & \frac {H_k}{\sqrt H}, \ \ \ \mathrm{k \times in\rightarrow\mathrm df} \le \mathrm{in\rightarrow fCutoff}, \\ & = & 0, \ \ \ \mathrm{k \times in\rightarrow df} > \mathrm{in\rightarrow fCutoff}. \end{eqnarray}
In addition, the 0th and Nyquist frequency components are also set to zero.
Definition in file LALInspiralWaveNormaliseLSO.c.
Go to the source code of this file.
| void LALInspiralWaveNormaliseLSO | ( | LALStatus * | status, |
| REAL4Vector * | filter, | ||
| REAL8 * | norm, | ||
| InspiralWaveNormaliseIn * | in | ||
| ) |
Definition at line 76 of file LALInspiralWaveNormaliseLSO.c.