Julianus Pfeuffer
Julianus Pfeuffer
Unless autowrap+msvc complained just about the potential type incompatibility, I would keep the extension and change the types there as well But you could try.
Did you test it on windows? Because our night lies still fail because of numpy
@timosachsenberg didn't you recently fix GitHub actions? Also pyopenms?
Specifically, we are missing: - Quality control (e.g. looking and browsing through extracted isotope traces in each found feature) - Parameter optimization (the noise threshold makes such a HUGE difference,...
Nice! Thanks a lot for trying and reporting. That was quick setting up their pipeline! How big are your files? I was using the benchmark dataset from https://www.ebi.ac.uk/metabolights/MTBLS736/files and FFMetabo...
> FeatureFinder and MetaboAdduct decharger are really fast, at least with my samples (which are quite complex). I have now run their workflow (actually using docker is a super good...
> The requantification step for OpenMS: FeatureFinderMetaboIdent took 6.44 s (wall), 5.57 s (CPU), 0.26 s (system), 5.31 s (user); Peak Memory Usage: 118 MB. > > I guess these...
> Hm all my files are max 40MB each. I m trying to download the ones you sent - any idea about their size? Unzipped as mzml they are around...
😆 what? which params? e.g. noise_threshold, snr, fwhm?
Right now, I have noise 80000 just to make my workflow run through :D But I also have some very loose RT settings, since I saw quite some small traces...