- Check the validity of a given template bank using a reference population of simulated CBC signals.
- Different approximants for templates and signals are supported -- Only aligned-spin templates supported at the moment -- Eccentric and precessing signals
- Handles large template bank and signal population with ease -- Parallelization using PyCBC workflows implemented via the Pegasus API -- Support for launching multiple workflows in parallel
Table of Contents
- Banksim for simulated signals from compact binary mergers
- Requirements
- Running the code * Generating the workflow - Submitting the workflow - Consolidating the results
For installation please visit the link https://pycbc.org/
Generate a template bank using either a lattice based or stochastic placement algorithm. For more details visit http://pycbc.org/pycbc/latest/html/tmpltbank.html
The template bank should be an HDF file and must contain the following parameters
| Parameter | Key |
|---|---|
| primary mass | mass1 |
| secondary mass | mass2 |
| z component of spin-1 | spin1z |
| z component of spin-2 | spin2z |
| eccentricity at a reference freq | eccentricity |
| mean anomaly parameter | long_asc_nodes |
| chirp time | tau0 |
| chirp time | tau3 |
The template bank must be sorted in increasing values of tau0
To generate the injections please visit http://pycbc.org/pycbc/latest/html/inference/examples/bbh.html
Injection parameters for all signals must be stored in a single HDF file. The injection file must contain the following parameters in addition to the template bank parameters
| Parameter | Key |
|---|---|
| x component of spin-1 | spin1x |
| y component of spin-1 | spin1y |
| x component of spin-2 | spin2x |
| y component of spin-2 | spin2y |
| eccentricity at a reference freq | eccentricity |
| mean anomaly parameter | long_asc_nodes |
| inclination of the orbit | inclination |
| polarization | polarization |
| distance to the source | distance |
| right ascention of the orbit | ra |
| source declination | dec |
Make sure the injection files contain the tau0 and tau3 parameters
Example PSD files can be generated via http://pycbc.org/pycbc/latest/html/psd.html Current support only for a PSD file and not ASD.
Once the template bank, injections and PSD files are ready, lastly, we need to set up the configuration file. An example_workflow.ini is provided with the code.
First change the path to the executable from [executables] section.
There are three important sections in the configuration
| Sections | Purpose |
|---|---|
| [Files] | Contains the full path to the input files |
| [Required] | Control the number of processes or workflows to split the analysis into. |
| [FF] | Arguments that are passed to the fitting-factor computation |
-
tau0_tolerance -- Reduce the number of templates for the match calculation by comparing the tau0s of the injection and templates; it can be expected that the best matching template will have a tau0 close to the injection's. To allow some mismatch, we implement a tolerance on this value. A higher tolerance will incorporate more templates in the analysis.
-
nsplits -- Split the match computation of one signal into these many procecesses.
-
nworkflow -- Split the injections into multiple workflows. A value of 1 is recommended unless the template bank and no. of injections are very large.
- approximant_tb and approximant_sg -- Approximants for the templates and signals respectively. Note only aligned spin templates are supported.
- HMs (Only for signals) -- Switch on-off the higher-modes of gravitaional wave. Provide 0 for no HMs and 1 to include HMs
- sampling_freq and sampling_rate -- Provide the values ensuring the Nyquist criterion requirements.
- detector -- Choose a single detector from ['H1', 'L1', 'V1']
- f_min -- Low frequency cutoff for the match computation
This primarily involves three steps -- generating the workflow, submitting the workflow and consolidating the results.
Navigate to the codebase directory and make subdirectory. This will be the top level directory for the analysis.
-
Open the gen.sh file and change the current working dir. The path will be the codebase directory in your local machine.
-
Change first and last -- corresponding to the indices of the injections analyzed from [first .... last] from the injection file.
-
Change the workflow_config
-
Execute the workflow generation using
./gen.sh sub_directory/
Navigate to the sub-directory and launch the submit.sh and provide the number of workflows (usually 1).
./submit.sh 1
After all the jobs are finished, launch the combine_FFs.py and provide number of injections and no. of splits for each injection.
python combine_FFs.py 10 10