Exemple #1
0
 def dw_stack(self, datadir, outdir = None, fskip=0, networks= [], channel='ZZ', vmin = 1., vmax = 5., Tmin = 5., Tmax = 150.,\
         prefer_c3_disp = True,  bfact_dw = 1., efact_dw = 1., snr_thresh = 10., ftan_type = 'DISPpmf2',\
         parallel = False,  nprocess=None, subsize=1000, verbose = True):
     """ stack direct wave interferogram
     =======================================================================================================
     ::: input parameters :::
     datadir         - directory including data
     outdir          - output directory
     channel         - channel pair for aftan analysis(e.g. 'ZZ', 'TT', 'ZR', 'RZ'...)
     vmin/vmax       - min/max group velocity to find max amplitude
     Tmin/Tmax       - min/max periods
     pers            - periods for stacking
     snr_thresh      - threshold SNR
     use_xcorr_aftan - use aftan results from xcorr as reference curve or not
     ftan_type       - ftan type of xcorr
     =======================================================================================================
     """
     if outdir is None:
         outdir  = datadir
     #---------------------------------
     # prepare data
     #---------------------------------
     print ('[%s] [DW_STACK] preparing C3 stack' %datetime.now().isoformat().split('.')[0])
     chan1                   = 'C3'+channel[0]
     chan2                   = 'C3'+channel[1]
     c3_lst                  = []
     for staid1 in self.waveforms.list():
         netcode1, stacode1      = staid1.split('.')
         with warnings.catch_warnings():
             warnings.simplefilter("ignore")
             tmppos1         = self.waveforms[staid1].coordinates
             stla1           = tmppos1['latitude']
             stlo1           = tmppos1['longitude']
         for staid2 in self.waveforms.list():
             netcode2, stacode2  = staid2.split('.')
             if staid1 >= staid2:
                 continue
             if len(networks) > 0:
                 if (not (netcode1 in networks)) and (not (netcode2 in networks)):
                     continue
             with warnings.catch_warnings():
                 warnings.simplefilter("ignore")
                 tmppos2         = self.waveforms[staid2].coordinates
                 stla2           = tmppos2['latitude']
                 stlo2           = tmppos2['longitude']
             # skip or not
             logfname    = datadir + '/logs_dw_stack/'+staid1+'/'+staid1+'_'+staid2+'.log'
             if os.path.isfile(logfname):
                 if fskip == 2:
                     continue
                 with open(logfname, 'r') as fid:
                     logflag = fid.readlines()[0].split()[0]
                 if (logflag == 'SUCCESS' or logflag == 'NODATA') and fskip == 1:
                     continue
                 if (logflag != 'FAILED') and fskip == -1: # debug
                     continue
             else:
                 if fskip == -1:
                     continue
             try:
                 subdset     = self.auxiliary_data[ftan_type][netcode1][stacode1][netcode2][stacode2][channel]
                 with warnings.catch_warnings():
                     warnings.simplefilter("ignore")
                     data        = subdset.data[()]
                     index       = subdset.parameters
                 Np              = int(index['Np'])
                 # reference curves
                 pers_ref        = data[index['To']][:Np]
                 phvel_ref       = data[index['C']][:Np]
                 if Np < 5:
                     print ('*** WARNING: Not enough I2 datapoints for: '+ staid1+'_'+staid2+'_'+channel)
                     phvel_ref   = []
                     pers_ref    = []
             except KeyError:
                 phvel_ref   = []
                 pers_ref    = []
                 # print ('!!! no reference phase velocity from I2: '+staid1+'_'+staid2)
             temp_c3_pair        = _c3_funcs.c3_pair(datadir = datadir, outdir = outdir, stacode1 = stacode1, netcode1 = netcode1,\
                 stla1 = stla1, stlo1 = stlo1,  stacode2 = stacode2, netcode2 = netcode2, stla2 = stla2, stlo2 = stlo2,\
                 channel = channel, vmin = vmin, vmax = vmax, Tmin = Tmin, Tmax = Tmax,  bfact_dw = bfact_dw, efact_dw = efact_dw,\
                 phvel_ref = phvel_ref, pers_ref = pers_ref, prefer_c3_disp = prefer_c3_disp)
             c3_lst.append(temp_c3_pair)
     #===============================
     # phase shift stack
     #===============================
     print ('[%s] [DW_STACK] computating... ' %datetime.now().isoformat().split('.')[0] )
     # parallelized run
     if parallel:
         #-----------------------------------------
         # Computing xcorr with multiprocessing
         #-----------------------------------------
         if len(c3_lst) > subsize:
             Nsub            = int(len(c3_lst)/subsize)
             for isub in range(Nsub):
                 print ('[%s] [DW_STACK] subset:' %datetime.now().isoformat().split('.')[0], isub, 'in', Nsub, 'sets')
                 cur_c3Lst   = c3_lst[isub*subsize:(isub+1)*subsize]
                 DW_STACK    = partial(_c3_funcs.direct_wave_phase_shift_stack_for_mp, verbose = verbose)
                 pool        = multiprocessing.Pool(processes=nprocess)
                 pool.map(DW_STACK, cur_c3Lst) #make our results with a map call
                 pool.close() #we are not adding any more processes
                 pool.join() #tell it to wait until all threads are done before going on
             cur_c3Lst       = c3_lst[(isub+1)*subsize:]
             DW_STACK        = partial(_c3_funcs.direct_wave_phase_shift_stack_for_mp, verbose = verbose)
             pool            = multiprocessing.Pool(processes=nprocess)
             pool.map(DW_STACK, cur_c3Lst) #make our results with a map call
             pool.close() #we are not adding any more processes
             pool.join() #tell it to wait until all threads are done before going on
         else:
             DW_STACK        = partial(_c3_funcs.direct_wave_phase_shift_stack_for_mp, verbose = verbose)
             pool            = multiprocessing.Pool(processes=nprocess)
             pool.map(DW_STACK, c3_lst) #make our results with a map call
             pool.close() #we are not adding any more processes
             pool.join() #tell it to wait until all threads are done before going on
     else:
         for ilst in range(len(c3_lst)):
             c3_lst[ilst].direct_wave_phase_shift_stack(verbose = verbose)     
     print ('[%s] [DW_STACK] ALL done' %datetime.now().isoformat().split('.')[0])
     return 
      
             
Exemple #2
0
 def dw_stack_disp(self, datadir, outdir = None, fskip = 0, networks= [], channel = 'ZZ', pers = [], vmin = 1.0, vmax = 4.5,
         snr_thresh = 10., Ntrace_min = 5, nfmin = 5, jump_thresh = 3., parallel = False, \
         nprocess=None, subsize=1000, verbose = False):
     """ stack dispersion results
     =======================================================================================================
     ::: input parameters :::
     datadir     - directory including data
     outdir      - output directory
     fskip       - skip upon dispersion output existence
                     0: overwrite
                     1: skip if success/nodata
                     2: skip upon log file existence
                     -1: debug purpose
     channel     - channel pair for aftan analysis(e.g. 'ZZ', 'TT', 'ZR', 'RZ'...)
     pers        - periods for stacking
     snr_thresh  - threshold SNR
     Ntrace_min  - minimum number of traces
     nfmin       - minimum frequency (period) data points for each trace
     jump_thresh - threshold value for jump detection
     
     parallel    - run the xcorr parallelly or not
     nprocess    - number of processes
     subsize     - subsize of processing list, use to prevent lock in multiprocessing process
     =======================================================================================================
     """
     if outdir is None:
         outdir  = datadir
     print ('[%s] [DW_STACK_DISP] start stacking direct c3 aftan results' %datetime.now().isoformat().split('.')[0])
     if len(pers) == 0:
         pers    = np.append( np.arange(18.)*2.+6., np.arange(4.)*5.+45.)
     c3_lst  = []
     for staid1 in self.waveforms.list():
         netcode1, stacode1      = staid1.split('.')
         with warnings.catch_warnings():
             warnings.simplefilter("ignore")
             tmppos1         = self.waveforms[staid1].coordinates
             stla1           = tmppos1['latitude']
             stlo1           = tmppos1['longitude']
         for staid2 in self.waveforms.list():
             netcode2, stacode2  = staid2.split('.')
             if staid1 >= staid2:
                 continue
             if len(networks) > 0:
                 if (not (netcode1 in networks)) and (not (netcode2 in networks)):
                     continue
             with warnings.catch_warnings():
                 warnings.simplefilter("ignore")
                 tmppos2         = self.waveforms[staid2].coordinates
                 stla2           = tmppos2['latitude']
                 stlo2           = tmppos2['longitude']
             # skip or not
             logfname    = datadir + '/logs_dw_stack_disp/'+staid1+'/'+staid1+'_'+staid2+'.log'
             if os.path.isfile(logfname):
                 if fskip == 2:
                     continue
                 with open(logfname, 'r') as fid:
                     logflag = fid.readlines()[0].split()[0]
                 if (logflag == 'SUCCESS' or logflag == 'NODATA') and fskip == 1:
                     continue
                 if (logflag != 'FAILED') and fskip == -1: # debug
                     continue
             else:
                 if fskip == -1:
                     continue
             c3_lst.append(_c3_funcs.c3_pair(datadir = datadir, outdir = outdir, stacode1 = stacode1, netcode1 = netcode1,\
                 stla1 = stla1, stlo1 = stlo1,  stacode2 = stacode2, netcode2 = netcode2, stla2 = stla2, stlo2 = stlo2, \
                 vmin = vmin, vmax = vmax, channel = channel, snr_thresh = snr_thresh, Ntrace_min = Ntrace_min,\
                 nfmin = nfmin, jump_thresh = jump_thresh))
     #===============================
     # direct wave interferometry
     #===============================
     print ('[%s] [DW_STACK_DISP] computating... ' %datetime.now().isoformat().split('.')[0] )
     # parallelized run
     if parallel:
         #-----------------------------------------
         # Computing xcorr with multiprocessing
         #-----------------------------------------
         if len(c3_lst) > subsize:
             Nsub            = int(len(c3_lst)/subsize)
             for isub in range(Nsub):
                 print ('[%s] [DW_STACK_DISP] subset:' %datetime.now().isoformat().split('.')[0], isub, 'in', Nsub, 'sets')
                 cur_c3Lst   = c3_lst[isub*subsize:(isub+1)*subsize]
                 AFTAN_STACK = partial(_c3_funcs.direct_wave_stack_disp_for_mp, verbose = verbose)
                 pool        = multiprocessing.Pool(processes=nprocess)
                 pool.map(AFTAN_STACK, cur_c3Lst) #make our results with a map call
                 pool.close() #we are not adding any more processes
                 pool.join() #tell it to wait until all threads are done before going on
             cur_c3Lst       = c3_lst[(isub+1)*subsize:]
             AFTAN_STACK     = partial(_c3_funcs.direct_wave_stack_disp_for_mp, verbose = verbose)
             pool            = multiprocessing.Pool(processes=nprocess)
             pool.map(AFTAN_STACK, cur_c3Lst) #make our results with a map call
             pool.close() #we are not adding any more processes
             pool.join() #tell it to wait until all threads are done before going on
         else:
             AFTAN_STACK     = partial(_c3_funcs.direct_wave_stack_disp_for_mp, verbose = verbose)
             pool            = multiprocessing.Pool(processes=nprocess)
             pool.map(AFTAN_STACK, c3_lst) #make our results with a map call
             pool.close() #we are not adding any more processes
             pool.join() #tell it to wait until all threads are done before going on
     else:
         for ilst in range(len(c3_lst)):
             c3_lst[ilst].direct_wave_stack_disp(verbose = verbose)     
     print ('[%s] [DW_STACK_DISP] all done' %datetime.now().isoformat().split('.')[0])
     return
Exemple #3
0
 def dw_interfere(self, datadir, outdir = None, networks= [], channel='ZZ', chan_types=['LH', 'BH', 'HH'], \
         alpha = 0.01, dthresh = 5., parallel=False, nprocess=None, subsize=1000, verbose=True, verbose2=False):
     """
     compute three station direct wave interferometry
     =================================================================================================================
     ::: input parameters :::
     datadir     - directory including data
     outdir      - output directory
     channel     - channel 
     chan_types  - types (also used as priorities) of channels (NOT implemented yet)
     alpha       - threshhold percentage to determine stationary phase zone
     dtresh      - threshhold difference in distance ( abs( abs(del_d_hyp) - del_d_ell) > dtresh )
     parallel    - run the xcorr parallelly or not
     nprocess    - number of processes
     subsize     - subsize of processing list, use to prevent lock in multiprocessing process
     =================================================================================================================
     """
     StationInv          = obspy.Inventory()        
     for staid in self.waveforms.list():
         with warnings.catch_warnings():
             warnings.simplefilter("ignore")
             StationInv  += self.waveforms[staid].StationXML
     if outdir is None:
         outdir  = datadir 
     #---------------------------------
     # prepare data
     #---------------------------------
     print ('[%s] [DW_INTERFERE] preparing for three station direct wave interferometry' %datetime.now().isoformat().split('.')[0])
     c3_lst  = []
     for staid1 in self.waveforms.list():
         netcode1, stacode1      = staid1.split('.')
         with warnings.catch_warnings():
             warnings.simplefilter("ignore")
             tmppos1         = self.waveforms[staid1].coordinates
             stla1           = tmppos1['latitude']
             stlo1           = tmppos1['longitude']
         for staid2 in self.waveforms.list():
             netcode2, stacode2  = staid2.split('.')
             if staid1 >= staid2:
                 continue
             if len(networks) > 0:
                 if (not (netcode1 in networks)) and (not (netcode2 in networks)):
                     continue
             with warnings.catch_warnings():
                 warnings.simplefilter("ignore")
                 tmppos2         = self.waveforms[staid2].coordinates
                 stla2           = tmppos2['latitude']
                 stlo2           = tmppos2['longitude']
             c3_lst.append(_c3_funcs.c3_pair(datadir = datadir, outdir = outdir, stacode1 = stacode1, netcode1 = netcode1,\
                 stla1 = stla1, stlo1 = stlo1,  stacode2 = stacode2, netcode2 = netcode2, stla2 = stla2, stlo2 = stlo2,\
                 channel = channel, chan_types = chan_types, StationInv = StationInv, alpha = alpha, \
                 dthresh = dthresh))
     #===============================
     # direct wave interferometry
     #===============================
     print ('[%s] [DW_INTERFERE] computating... ' %datetime.now().isoformat().split('.')[0] )
     # parallelized run
     if parallel:
         #-----------------------------------------
         # Computing xcorr with multiprocessing
         #-----------------------------------------
         if len(c3_lst) > subsize:
             Nsub            = int(len(c3_lst)/subsize)
             for isub in range(Nsub):
                 print ('[%s] [DW_INTERFERE] subset:' %datetime.now().isoformat().split('.')[0], isub, 'in', Nsub, 'sets')
                 cur_c3Lst   = c3_lst[isub*subsize:(isub+1)*subsize]
                 CCUBE       = partial(_c3_funcs.direct_wave_interfere_for_mp, verbose = verbose, verbose2 = verbose2)
                 pool        = multiprocessing.Pool(processes=nprocess)
                 pool.map(CCUBE, cur_c3Lst) #make our results with a map call
                 pool.close() #we are not adding any more processes
                 pool.join() #tell it to wait until all threads are done before going on
             cur_c3Lst       = c3_lst[(isub+1)*subsize:]
             CCUBE           = partial(_c3_funcs.direct_wave_interfere_for_mp, verbose = verbose, verbose2 = verbose2)
             pool            = multiprocessing.Pool(processes=nprocess)
             pool.map(CCUBE, cur_c3Lst) #make our results with a map call
             pool.close() #we are not adding any more processes
             pool.join() #tell it to wait until all threads are done before going on
         else:
             CCUBE           = partial(_c3_funcs.direct_wave_interfere_for_mp, verbose = verbose, verbose2 = verbose2)
             pool            = multiprocessing.Pool(processes=nprocess)
             pool.map(CCUBE, c3_lst) #make our results with a map call
             pool.close() #we are not adding any more processes
             pool.join() #tell it to wait until all threads are done before going on
     else:
         Nsuccess    = 0
         Nnodata     = 0
         for ilst in range(len(c3_lst)):
             if c3_lst[ilst].direct_wave_interfere(verbose = verbose, verbose2 = verbose2) > 0:
                 Nsuccess+= 1
             else:
                 Nnodata += 1
         print ('[%s] [DW_INTERFERE] computation ALL done: success/nodata: %d/%d' %(datetime.now().isoformat().split('.')[0], Nsuccess, Nnodata))
     return
Exemple #4
0
 def dw_aftan(self, datadir, prephdir, fskip = 0, networks= [], channel='ZZ', outdir = None, inftan = pyaftan.InputFtanParam(),\
         basic1=True, basic2=True, pmf1=True, pmf2=True, verbose = True, f77=True, pfx='DISP', parallel = False, \
         nprocess=None, subsize=1000):
     """direct wave interferometry aftan
     =================================================================================================================
     datadir     - directory including data
     prephdir    - directory for predicted phase velocity dispersion curve
     fskip       - skip upon dispersion output existence
                     0: overwrite
                     1: skip if success/nodata
                     2: skip upon log file existence
                     -1: debug purpose
     channel     - channel pair for aftan analysis(e.g. 'ZZ', 'TT', 'ZR', 'RZ'...)
     outdir      - directory for output disp txt files (default = None, no txt output)
     inftan      - input aftan parameters
     basic1      - save basic aftan results or not
     basic2      - save basic aftan results(with jump correction) or not
     pmf1        - save pmf aftan results or not
     pmf2        - save pmf aftan results(with jump correction) or not
     f77         - use aftanf77 or not
     pfx         - prefix for output txt DISP files
     parallel    - run the xcorr parallelly or not
     nprocess    - number of processes
     subsize     - subsize of processing list, use to prevent lock in multiprocessing process
     =================================================================================================================
     """
     if outdir is None:
         outdir  = datadir
     #---------------------------------
     # prepare data
     #---------------------------------
     print ('[%s] [DW_AFTAN] preparing for three station direct wave aftan' %datetime.now().isoformat().split('.')[0])
     c3_lst  = []
     for staid1 in self.waveforms.list():
         netcode1, stacode1      = staid1.split('.')
         with warnings.catch_warnings():
             warnings.simplefilter("ignore")
             tmppos1         = self.waveforms[staid1].coordinates
             stla1           = tmppos1['latitude']
             stlo1           = tmppos1['longitude']
         for staid2 in self.waveforms.list():
             netcode2, stacode2  = staid2.split('.')
             if staid1 >= staid2:
                 continue
             if len(networks) > 0:
                 if (not (netcode1 in networks)) and (not (netcode2 in networks)):
                     continue
             with warnings.catch_warnings():
                 warnings.simplefilter("ignore")
                 tmppos2         = self.waveforms[staid2].coordinates
                 stla2           = tmppos2['latitude']
                 stlo2           = tmppos2['longitude']
             # skip or not
             logfname    = datadir + '/logs_dw_aftan/'+staid1+'/'+staid1+'_'+staid2+'.log'
             if os.path.isfile(logfname):
                 if fskip == 2:
                     continue
                 with open(logfname, 'r') as fid:
                     logflag = fid.readlines()[0].split()[0]
                 if (logflag == 'SUCCESS' or logflag == 'NODATA') and fskip == 1:
                     continue
                 if (logflag != 'FAILED') and fskip == -1: # debug
                     continue
             else:
                 if fskip == -1:
                     continue
             c3_lst.append(_c3_funcs.c3_pair(datadir = datadir, outdir = outdir, stacode1 = stacode1, netcode1 = netcode1,\
                 stla1 = stla1, stlo1 = stlo1,  stacode2 = stacode2, netcode2 = netcode2, stla2 = stla2, stlo2 = stlo2,\
                 channel = channel, inftan = inftan, basic1=basic1, basic2=basic2, pmf1=pmf1, pmf2=pmf2, f77=f77, prephdir = prephdir))
     #===============================
     # direct wave interferometry
     #===============================
     print ('[%s] [DW_AFTAN] computating... ' %datetime.now().isoformat().split('.')[0] )
     # parallelized run
     if parallel:
         #-----------------------------------------
         # Computing xcorr with multiprocessing
         #-----------------------------------------
         if len(c3_lst) > subsize:
             Nsub            = int(len(c3_lst)/subsize)
             for isub in range(Nsub):
                 print ('[%s] [DW_AFTAN] subset:' %datetime.now().isoformat().split('.')[0], isub, 'in', Nsub, 'sets')
                 cur_c3Lst   = c3_lst[isub*subsize:(isub+1)*subsize]
                 AFTAN       = partial(_c3_funcs.direct_wave_aftan_for_mp, verbose = verbose)
                 pool        = multiprocessing.Pool(processes=nprocess)
                 pool.map(AFTAN, cur_c3Lst) #make our results with a map call
                 pool.close() #we are not adding any more processes
                 pool.join() #tell it to wait until all threads are done before going on
             cur_c3Lst       = c3_lst[(isub+1)*subsize:]
             AFTAN           = partial(_c3_funcs.direct_wave_aftan_for_mp, verbose = verbose)
             pool            = multiprocessing.Pool(processes=nprocess)
             pool.map(AFTAN, cur_c3Lst) #make our results with a map call
             pool.close() #we are not adding any more processes
             pool.join() #tell it to wait until all threads are done before going on
         else:
             AFTAN           = partial(_c3_funcs.direct_wave_aftan_for_mp, verbose = verbose)
             pool            = multiprocessing.Pool(processes=nprocess)
             pool.map(AFTAN, c3_lst) #make our results with a map call
             pool.close() #we are not adding any more processes
             pool.join() #tell it to wait until all threads are done before going on
     else:
         Nsuccess    = 0
         Nnodata     = 0
         for ilst in range(len(c3_lst)):
             c3_lst[ilst].direct_wave_aftan(verbose = verbose) 
     print ('[%s] [DW_AFTAN] computation ALL done' %datetime.now().isoformat().split('.')[0])
     return