Tom Winder
Nice - I think this is a good stop-gap solution. Couple of things: * We will need to do something about the required Travis check - possibly can be overridden....
Slight curveball: I also tried installing from source on macOS, and have different behaviour compared to the previous setup.py : ```diff - gcc -Wno-unused-result -Wsign-compare -Wunreachable-code -DNDEBUG -g -fwrapv -O3...
Testing shows that the changed behaviour is due to a different (seemingly now more accurate) platform test in the updated `setup.py` -- looking now at the output of `platform.system()` rather...
Re-opened pending addition of multi-processing to trigger.
I'd class this as a bug still, because it's important to be able to try a high threshold and then inspect the plot to guide your choice of a lower...
Tim can you test the fix I have applied at f03b290, please? I have tested it locally and seems to fix the issue.
As mentioned in the linked issue (#93) a lot of this code is moving in a change that is currently underway. These changes would therefore have to be largely re-written...
1. Ok, good. 2. Again, please make suggestions like this as issues / directly to us - the feedback is super useful, but the most efficient way to improve the...
Yeh, to re-emphasise: the next update to development (in review atm) will significantly change how waveform data is handled internally. So we have an opportunity to take stock and decide...
I am testing 0.0.9 against lut.read_3d_nlloc_lut() and lut.compute_1d_vmodel() to see if there's anything obviously wrong...