Running the search workflows

We use large-scale PyCBC search workflows to evaluate the sensitivity of our search method in finding precessing signals in real data. This step is computationally expensive and does require a cluster running condor.

Code base

We used a modified branch of PyCBC to do this step. This is the same branch as used for other stages later on. The code we used is here

https://github.com/spxiwh/pycbc/tree/tha_development_work_rebased_again

We will aim to merge these changes back into the main PyCBC branch, but this was the branch used when creating the banks used in the paper. We are using some modified executables in this step. These have the suffix _tha. We will consider whether these can be merged with their original executables.

Files needed for the search workflows

There are a number of files needed for the search workflows. Here are all of them (for all of the searches that are needed):

Running the workflows

The workflow can be submitted to condor using:

pycbc_make_coinc_search_workflow --workflow-name search --output-dir DIR_NAME --config-files CONFIG_FILE --submit-now

This requires choosing an output directory, and one of the configuration files. There are 6 configuration files here, corresponding to the 6 searches presented in the paper.

Some things to look out for in the configuration files

The configuration files will likely need some minor edits to work in your setup. Specifically:

Look out for the injections-pregenerated-file options. Make sure these point to the files downloaded above.
The [executables] section links to the various executables. This would need to point to your versions.
The [results_page] section contains a directory for where results will appear. This will need updating.
The statistics-files option in the [coinc] section will need to point to the files downloaded above (note that a different file is used for the precessing vs aligned-spin searches here).
The tmpltbank-pregenerated-bank option in the [workflow-tmpltbank] section needs to point to the precessing, or aligned-spin, bank either generated following earlier instructions or downloaded from the bank generation section.
Many sections: [workflow-datafind] and [workflow-segments] and [workflow] -> XX-channel-name and channel-name/frame-type in other sections are setup to access data using LIGO resources. These data are the same as those available from GWOSC, but the PyCBC tools need some changes to use GWOSC to access them. We refer the reader to The 4-OGC paper data release for details on how to do this on O3 data.

Outputs

These workflows will generate the injection results used in our plotting codes. We make available these outputs here.

Note that Figure 7 and 8 use additional files than this. Those files are very large (in some cases > 5GB), and therefore not easy to make available on this platform. We are happy to share these files on request, and will try to identify a suitable platform to host them. The files can also be regenerated following these instructions and running the search.