1.8 KiB
Signal injection and testing for GREENBURST
These scripts are developed to:
- Create and inject bursts of varying DM and width to existing FILTERBANK files (signalgen.py), and
- Analyze the results of running the resulting files through the GREENBURST pipeline (testanalysis.py).
Environment
The inject_env.yml file contains the enviroment declaration necessary to run these scripts. On the GREENBURST machine, this enviroment is accessed via conda activate inject.
Running
Signal Generation
Signalgen has 2 modes - generation and plotting. Generation is the default, while plotting is accessed via the -p flag.
Files may be provided in three ways:
- Directly and one at a time, via
-f - Through a file containing a list of file paths, via
-l - Contained in a directory via
-D, where Signalgen will act on all.filfiles in the directory.
The injected signals can be of a specified DM (-d) and/or width (-w), and the number of samples taken by the will package can be adjusted as well (-n).
For bulk processing, the -r flag should be used to randomize the pulse parameters and output the injected pulse infomation to a timestamped file.
By default, the resulting files will be output to ./out/, but a different directory may be specified with -o. Note that -p will always output the plots into the same directory as the source .fil file.
Test Analysis
After running the GREENBURST pipeline using the -t flag with stage 3, a plotOut.txt file will be output (by default, to /ldata/dev_tests/repro/). Provide the timestamped file from Signalgen and this plotOut.txt file to Testanalysis by editing the paths at the top of the script, and run the file. Information will be output to the log, and plots will be output to ./out/.