SNR calculation with IMRPhenomHM - JAMSADIQ/LISAPopulationAnalysis GitHub Wiki

I used posterior samples for each events for all parameters, intrinsic (M, q, z, Xi1, Xi2), extrinsic(No MonteCarlo) and use script to compute the optimal SNR using lisabeta. The waveform model is IMRPhenomHM

Large mass ratio and high redshift posteriors

There are some events which have posterior samples with large mass ratio (q > 100) and redshift (z > 30) and for which SNR values are very small. These events are following.

• catalog_1_yrs.txt.92.87.h5
• catalog_1_yrs.txt.92.195.h5
• catalog_1_yrs.txt.92.332.h5
• catalog_1_yrs.txt.92.377.h5

The plots for SNR as function of mass ratio and redshift are here

Posteriors with nan SNR

There are few samples in events below where SNR output was nan. I add the index of that parameter from .h5 file. I notices that if I change "timetomerger_max" in my waveform_params_mbhb dictionary (from default value 4 to value 1), these parameters give reasonable SNR value but than some other parameter values give nan and I did not understand that part.

• catalog_1_yrs.txt.92.203.h5 , index 6600
• catalog_1_yrs.txt.92.15.h5 , index 27132
• catalog_1_yrs.txt.92.172.h5 , index 6657
• catalog_1_yrs.txt.92.119.h5 , index 11695

The one common features in SNRs on samples for these files is that there are few SNR values which are too high compared to the most other samples and one event plot is here to check it.