Header file for the signal injection functions for the parameter estimation code for known pulsar searches using the nested sampling algorithm. More...
Prototypes | |
void | inject_signal (LALInferenceRunState *runState) |
Inject a simulated signal into the data. More... | |
REAL8 | calculate_time_domain_snr (LALInferenceIFOData *data, LALInferenceIFOModel *ifo_model) |
Calculates the optimal matched filter signal-to-noise ratio for a given signal. More... | |
void | get_loudest_snr (LALInferenceRunState *runState) |
Get the signal-to-noise ratio of the maximum likelihood signal. More... | |
Header file for the signal injection functions for the parameter estimation code for known pulsar searches using the nested sampling algorithm.
Definition in file ppe_inject.h.
Go to the source code of this file.
void inject_signal | ( | LALInferenceRunState * | runState | ) |
Inject a simulated signal into the data.
This function will create an simulated signal (of the required model) to inject into the data from multiple detectors. The parameters of the signal to be injected must be specified in a TEMPO-stype .par file given with the inject-file
command line argument. The parameters do not have to be the same as those in the .par file controlling the analysis (although should ideally contain a signal within the bandwidth of the data).
If a signal of a specific signal-to-noise ratio is required then the scale-snr
command line argument can be used to give the multi-detector SNR to which the signal needs to be scaled.
The injected signal can be output if inject-output
is set. Two files will be output: one containing the signal only, and one containing the signal plus noise. These will both be in the format of a standard data input file. The files will have names given by the inject-output
value, with a prefix of the detector name, and a suffix of of _signal_only
, respectively.
runState | [in] the program information structure |
Definition at line 47 of file ppe_inject.c.
REAL8 calculate_time_domain_snr | ( | LALInferenceIFOData * | data, |
LALInferenceIFOModel * | ifo_model | ||
) |
Calculates the optimal matched filter signal-to-noise ratio for a given signal.
This function calculates the optimal matched filter signal-to-noise ratio (SNR) of a given signal model for a set of detector data via:
\[ \rho = \sqrt{\sum_{i=1}^N \frac{d_i^2}{\sigma^2}}, \]
where \( \{d\} \) is a time series of data, and \( \sigma^2 \) is its variance. As the data and model used here are complex the real and imaginary SNRs are added in quadrature to give the total SNR.
The data variance \( \sigma^2 \) is calculated on data that has had the running median subtracted in order to remove any underlying trends (e.g. caused by a string signal). The variance is assumed constant over segments given in the chunkLength
vector and the SNR from each segment is added in quadrature.
data | [in] A data pointer containing detector data and the signal model |
ifo_model | [in] A model structure containing detector parameters and buffers |
Definition at line 503 of file ppe_inject.c.
void get_loudest_snr | ( | LALInferenceRunState * | runState | ) |
Get the signal-to-noise ratio of the maximum likelihood signal.
The function uses the signal with the highest likelihood (which will be the final point in the live points array) and calculates the optimal signal-to-noise ratio (SNR) for it. This is output to a file based on the outfile
value, but with _SNR
appended to it. For multiple detector, and/or models with multiple data sets, the individual detector/data set SNR values will be output, with the final value being the multi-detector SNR. If a fake signal has been injected into the data this file will already contain the optimal SNR of the true signal.
runState | [in] The analysis information structure |
Definition at line 647 of file ppe_inject.c.