Join GitHub today
GitHub is home to over 31 million developers working together to host and review code, manage projects, and build software together.Sign up
Nordic high accuracy #2351
What does this PR do?
Adds support for high(er)-precision timing in Nordic pick files (closes #2348) and add support for long phase names - from the seisan manual:
Why was it initiated? Any relevant Issues?
Seconds can run in columns 23-28 - I misread the indexing and was reading columns 23-28 rather than 22-28. #2348 pointed out that this is used when seisan is run in "high-accuracy" mode.
Note, pick times are now written out by default in high-accuracy from obspy now. I need to confirm that seisan is always happy with this, if not then a flag option for high-accuracy should be implemented.
This was referenced
Mar 11, 2019
I checked whether SEISAN was happy reading high-accuracy phase-picks when not run in high-accuracy mode, and it seems to be fine. Nevertheless a flag has been added for writing to turn off high accuracy mode (which is also tested).
circleci pep8 tests are not related to files I have changed for this:
and it looks like appveyor is failing on response removal things that are unrelated.
I think this is ready for review...
d-chambers left a comment
This looks good. It looks like you have comprehensively tested the new functionality.
One thing that might help the maintainability would be to move some of the utility functions (like string converters) into a obspy.io.seisan.util module as is done in many of the other IO modules. Core is getting rather long, and there has been some talk of trying to unify the io utils at some point in time.
Docs built fine, the only thing I changed is in the
Test fails on appveyor are for trace comparison after response removal, the tests that fail are:
Looks like the version of flake8 installed on circleci has been bumped from 3.6.0 to 3.7.7, which is resulting in a lot of import naming related complaints. Happy to trudge through marking those to be ignored, or changing them today if it helps (in another PR)?
The travis fail looked like a network issue, I have restarted it.
Good to go @d-chambers ?