OWC_3 --> pyOWC transition

[dirkslawinski]

HiHo,
so @kamwal has asked me to give pyOWC a go and see why it falls over for some. Argo Australia has been using OWC_3 for awhile now ... with some localized adds/fixes. We did get pyOWC installed-ish when it 1st came out but had not given it a true test. Now we will.

I note that unlike the Matlab version the "ow_config.txt" info is hardwired in /argodmqc_owc/pyowc/configuration.py. Was this a conscious decision to template the dictionary in the object which you then update manually or was there scope to read a local file file, say in json, at a later time?

[gmaze]

This is not a pythonic design and probably a choice that should change in the future

[dirkslawinski]

OK, my local process loads a JSON file to overwrite the templated info. Just as a suggestion for a future path.

Next step is to sort the numpy version issue.