Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Repo info
Activity
    Thomas Lecocq
    @ThomasLecocq
    Hey Sven (I guess?) Yep for 3 it's a change i made after writing the instructions... For the empty GUI that's odd... Are there any error message in the console?
    schipp
    @schipp
    Traceback (most recent call last):
      File "/Users/sven/anaconda3/envs/cargese19/lib/python3.7/site-packages/matplotlib/cbook/__init__.py", line 216, in process
        func(*args, **kwargs)
      File "/Users/sven/anaconda3/envs/cargese19/lib/python3.7/site-packages/msnoise_tomo/iftan.py", line 294, in onclick
        (event.button, event.x, event.y, event.xdata, event.ydata))
    TypeError: must be real number, not NoneType
    Also, I'm still having issues with SDS.zip. Redownloaded twice now but it says archive is "incomplete" and extracted only 2.15GB (zip is 6.11GB).
    Thomas Lecocq
    @ThomasLecocq
    Did you download it to a disk or a usb stick? Strange....
    For the error, seems the 'process' thingy is called when it shouldn't
    schipp
    @schipp
    Downloaded via filezilla straight to disk, it's 6113183074 bytes . I'd say let's try later via usb stick? Let me know, if I can do anything to get the GUI working. Gotta get going now, see you later
    Thomas Lecocq
    @ThomasLecocq
    No, to disk should be fine
    We'll check today or tomorrow morning for the gui
    Thomas Lecocq
    @ThomasLecocq
    could you send me (by email) the content of conda list ?
    Thomas Lecocq
    @ThomasLecocq
    does resizing the window change anything ? are you using a "dark theme" or something on Mac ?
    (what version of OsX is that?)
    schipp
    @schipp
    dark mode on or off makes no difference. resizing results in the window getting black. behaviour is the same with dark mode on/off. I also just realized that the error pasted above occurs when i click into the window. it's macos mojave
    Thomas Lecocq
    @ThomasLecocq
    so before you click into it, it's showing the menus/boxes ?
    (I read stuff about Mojave, we'll solve that tomorrow , haven't checked the workarounds)
    Thomas Lecocq
    @ThomasLecocq
    @schipp can you try removing tk and install this version: https://anaconda.org/conda-forge/tk/8.6.9/download/osx-64/tk-8.6.9-ha441bb4_1000.tar.bz2
    should also work by conda remove tk & conda install -c conda-forge tk=8.6.9=ha441bb4_1000
    François Lavoué
    @lavouef_gitlab
    Hi Thomas, I had the same issue with the GUI on MacOS 10.14, and reinstalling the environment with tk=8.6.9=ha441bb4_1000 seems to do the trick. Thanks!
    conda create -n cargese19_v2 -c conda-forge python=3.7 obspy=1.1.1 msnoise shapely pyproj tk=8.6.9=ha441bb4_1000
    NB: conda remove tk & conda install -c conda-forge tk=8.6.9=ha441bb4_1000 looks like it wants to remove everything but tk, so I didn't do it this way. Maybe it works too.
    Thomas Lecocq
    @ThomasLecocq
    Excellent! Thanks for checking, good it works
    please spread the word :-)
    François Lavoué
    @lavouef_gitlab
    Yes... except I'm running in another issue now: I can't open any folder, I get the following error
    $ Exception in Tkinter callback Traceback (most recent call last): File "//anaconda3/envs/cargese19_v2/lib/python3.7/tkinter/__init__.py", line 1705, in __call__ return self.func(*args) File "//anaconda3/envs/cargese19_v2/lib/python3.7/site-packages/msnoise_tomo/iftan.py", line 60, in load_dir cb_val.set(files[0]) IndexError: list index out of range
    $ Exception in Tkinter callback
    Traceback (most recent call last):
    File "//anaconda3/envs/cargese19_v2/lib/python3.7/tkinter/init.py", line 1705, in call
    return self.func(*args)
    File "//anaconda3/envs/cargese19_v2/lib/python3.7/site-packages/msnoise_tomo/iftan.py", line 60, in load_dir
    cb_val.set(files[0])
    IndexError: list index out of range
    (more readable this way)
    titi-dev
    @titi-dev
    hii sir. As per the manual I obtain daily cross-correlation stack in the stack folder.Can it be possible to obtain only one stacked trace per station pair by MSnoise instead of daily stack.if so how can i do it please suggest.
    Thomas Lecocq
    @ThomasLecocq
    @titi-dev : read the documentation or here above, I already replied to this
    @lavouef_gitlab and the gui still work or doesn't even start ?
    schipp
    @schipp
    @ThomasLecocq GUI working now. For me, opening the folder and computing the FTAN works, too
    Thomas Lecocq
    @ThomasLecocq
    then it's fiiiiiine :)
    titi-dev
    @titi-dev
    @ThomasLecocq sir it only produce daily stack as indicated by the sentence "CCF are stacked to produce a daily cross-correlation function".But I donot need daily stack instead only single stacked trace for every pair of station like what hermann's seismological noise code does. is the mov_stack does this thing. I am confused.
    Thomas Lecocq
    @ThomasLecocq
    Yes , there other steps and documentation pages to read... Proceed to the next step... Like 'the ref stack'
    François Lavoué
    @lavouef_gitlab
    Yes, everything is working now, thanks. Gosia just explained me that we had to open a folder that contains SAC file.
    Thomas Lecocq
    @ThomasLecocq
    amazing :)
    titi-dev
    @titi-dev
    @ThomasLecocq Sir i need to stack 60 days cross-correlated data to produce the single stacked trace for this i set mov_stack:60 and ref_begin =2018-01-01 and refend=2018-03-01 but its producing same thing only i.e. everyday stack in STACK directory.It doesnot even produce REF or DAYS??? directory.
    Thomas Lecocq
    @ThomasLecocq
    Pffffff msnoise stack -r ... Come on...
    titi-dev
    @titi-dev
    @ThomasLecocq yes it produced DAYS_060 directory but still inside DAYS_060 directory days stacks produced file names like 2018-01-01.SAC, 2018-01-02.SAC, 2018-01-03.SAC, 2018-01-04.SAC etc but it doesnot produce single stacked trace.
    titi-dev
    @titi-dev
    @ThomasLecocq it worked now.Sir i have a silly doubt. when i cross correlate the traces i am getting the shape of the traces like autocorrelagram where highest peak is just above zero lag.However i am not getting causal and acausal part of the cross correlogram.Am i doing wrong or some parameters i have to set properly inorder to get the stacked crosscorrelogram with causal and acausal part.please suggest
    Thomas Lecocq
    @ThomasLecocq
    depends on the filters used, the distance between stations etc... lots of parameters, yours to investigate
    titi-dev
    @titi-dev
    Now my question is how to obtain empirical greens function? Are the files inside the REF directory are the greens function or some more calculation i have to do.
    Thomas Lecocq
    @ThomasLecocq
    Guess
    And read
    titi-dev
    @titi-dev
    I think files inside REF directory are Empirical Green's function
    titi-dev
    @titi-dev
    @ThomasLecocq sir Is my answer right....
    titi-dev
    @titi-dev
    @ThomasLecocq Sir i red the Bensen et al.,2007 paper.In one place they have written that "After the daily cross-correlations have been computed and stacked,
    the resulting waveform is an estimated Green function" .So I think files inside REF directory are Empirical Green's function only.
    titi-dev
    @titi-dev
    hii...i want to use a bandpass filter how to set it in MSnoise config....please suggest.i didnot get any idea in manual
    Thomas Lecocq
    @ThomasLecocq
    @titi-dev i am sorry but i will not answer you anymore. I spent so long writing the paper and the documentation, your questions have ALL simple answers in the documentation. All. Good luck with your work.
    titi-dev
    @titi-dev
    @ThomasLecocq ok sir Thanks.
    titi-dev
    @titi-dev
    @ThomasLecocq plz donot mind sir if you get angry then how do i run MSnoise.One doubt i have....whenever i do noise cross corrlation using MSnoise there is Asymmetry of Noise Correlation Functions . How MSnoise overcome this problem....Does it compute the average of the Noise Correlation Function(which is not in the manual).
    Thomas Lecocq
    @ThomasLecocq
    nature is not theory... so non-symetrical noise correlation functions... is ... natural... because sources/scatterers are not evenly distributed around your network. There are papers discussing if (and how) those asymetrical CCF can be used for tomography.