Questions on usage

Created by: keflavich

I'm trying to run frocc on some ALMA data, since I can't get tclean's native parallel mode to work.

I have a lot of setup questions that I don't see in the docs and can't quite figure out by inspecting the source code.

I'm running commands like this one:

PYTHONPATH= frocc --createConfig --runScripts '["cube_split.py", "cube_tclean.py", "cube_buildcube.py", "cube_ior_flagging.py", "cube_average_map.py", "cube_report.py", "cube_cleanup.py"]' --slurmDefaultHeader "{'partition': '', 'account': 'adamginsburg', 'qos': 'adamginsburg-b', 'job-name': 'frocc_a35'}" --inputMS uid___A002_Xfdf45f_X38ff_target_spw35.ms,uid___A002_Xfdf45f_X4f4_target_spw35.ms,uid___A002_Xfdf45f_X38ff_target_spw35.ms --createScripts --start --freqRanges "[99550.85757-101484.5705195]"

First, I thought it was a bit weird that cube_split wasn't included in the default list - I had to add it.

Then, I changed wsclean to tclean.

I've been running into a wide variety of problems:

• frocc seems to overwrite the .frocc_default_config.template file on each run rather than respecting the local one
• tclean fails with errors like:

cube_tclean-64461335-1.err
25:    listing = glob(f"{conf.env.dirVis}/*{conf.env.markerChannel}*")
26:AttributeError: 'NoneType' object has no attribute 'dirVis'

There are a few other issues I've managed to debug by digging into the source code (e.g., the frequency range has to be specified in MHz), but it would be helpful to have a little more guidance on the treatment of config files.