Hi,
I was recently suggested the POWER ( https://git.ncsa.illinois.edu/elihu/Gravitational_Waveform_Extractor/) code for some GW extrapolation work I was doing by Gabriele Bozzola but I believe I found some bugs and he suggested I send an email here.
Running the script as is on the GW150914 gallery example gives me results significantly different from what the plots on the ET website show. I found replacing where the code currently fits the strain amplitude and phase to a polynomial expression with numpy polyfit seems to fix the main problem.
I also noticed in the getCutoffFrequencyFromTwoPuncturesBBH function, lines 341 - 347 have an incorrect equation for angular momentum, it looks like a typo between position and momentum. Right now its position * momentum - momentum^2.
I'm not sure if this code is being actively used, but I saw there were git branches for the latest ET release.
Thanks, Anuj Kankani
Hello Anuj,
Thank you or your email.
Those sound like bugs indeed. Could you create an ET ticket at
https://bitbucket.org/einsteintoolkit/tickets/issues/new
and I will try and see if I can incorporate the fixes you suggested quickly.
Can you provide your replacement fitting code that you refer to in:
I found replacing where the code currently fits the strain amplitude and phase to a polynomial expression with numpy polyfit seems to fix the main problem.
Yours, Roland
Hi,
I was recently suggested the POWER ( https://git.ncsa.illinois.edu/elihu/Gravitational_Waveform_Extractor/) code for some GW extrapolation work I was doing by Gabriele Bozzola but I believe I found some bugs and he suggested I send an email here.
Running the script as is on the GW150914 gallery example gives me results significantly different from what the plots on the ET website show. I found replacing where the code currently fits the strain amplitude and phase to a polynomial expression with numpy polyfit seems to fix the main problem.
I also noticed in the getCutoffFrequencyFromTwoPuncturesBBH function, lines 341 - 347 have an incorrect equation for angular momentum, it looks like a typo between position and momentum. Right now its position * momentum - momentum^2.
I'm not sure if this code is being actively used, but I saw there were git branches for the latest ET release.
Thanks, Anuj Kankani
users@lists.einsteintoolkit.org