def merger(noise_trace, st_events_poisson, samp_rate, delta):
""" Merges the noise with the events """
# Creates the stream with the noise and events
newnoise_t_formerging = noise_trace.slice(0, (st_events_poisson[-1].stats.endtime)+100)
seis = Stream()
seis += newnoise_t_formerging
seis += st_events_poisson
figs = plt.figure(figsize=(14,5))
seis.plot(fig = figs)
figs.suptitle("Noise and events", fontsize = 12)
ylab=figs.text(0.05, 0.5, 'Amplitude of signal', va='center',
rotation='vertical', fontsize=12)
figs.text(0.5, 0, 'Timestamp', ha='center', fontsize=12)
seis.merge()
seis += st_events_poisson
seis.merge(method = 1, interpolation_samples=-1)
figs = plt.figure(figsize=(14,5))
seis.plot(fig = figs)
plt.title("Synthetic seismogram", fontsize = 12)
ylab=figs.text(0.05, 0.5, 'Amplitude of signal', va='center',
rotation='vertical', fontsize=12)
figs.text(0.5, 0, 'Timestamp', ha='center', fontsize=12)
syn_seis = seis[0]
syn_seis.stats.sampling_rate = samp_rate
syn_seis.stats.delta = delta
return syn_seis
def _read_filepattern(fpattern, starttime, endtime, trim):
"""Read a stream from files whose names match a given pattern.
"""
flist = glob.glob(fpattern)
starttimes = []
endtimes = []
# first only read the header information
for fname in flist:
st = read(fname,headonly=True)
starttimes.append(st[0].stats.starttime.datetime)
endtimes.append(st[-1].stats.endtime.datetime)
# now read the stream from the files that contain the period
st = Stream()
for ind,fname in enumerate(flist):
if (starttimes[ind] < endtime) and (endtimes[ind] > starttime):
if trim:
st += read(fname,starttime=UTCDateTime(starttime),endtime=UTCDateTime(endtime))
else:
st += read(fname)
try:
st.merge()
except:
print "Error merging traces for requested period!"
st = Stream()
return st
def _do_select_internal(dev_id: str,
start_time: datetime.datetime,
end_time: datetime.datetime,
do_merge: bool = True) -> Stream:
"""
Retrieve stream from database
Constraint:
"""
# start_time, end_time = start_time.utcnow(), end_time.utcnow()
st = start_time.strftime('%Y-%m-%dT%H:%M:%S')
et = end_time.strftime('%Y-%m-%dT%H:%M:%S')
print(st, et)
# Build query
qry = (f"SELECT time_record_start, contents from skt_sensor "
f"WHERE sensor_id='{dev_id}' AND "
f"time_bucket='{start_time.date()}' AND "
f"time_record_start >= '{st}' AND "
f"time_record_start <= '{et}'")
result_set = session.execute(qry)
# Create empty stream
s = Stream()
# For each stream in result_set
for r in result_set:
target = zlib.decompress(r.contents)
bio = BytesIO(target)
s += obspy.read(bio)
# No result in result_set
if len(s) == 0:
return None
# Fix network code
for t in s:
t.stats.network = 'SK'
if do_merge:
s.merge(fill_value='latest')
return s
def _read_filepattern(fpattern, starttime, endtime, trim):
"""Read a stream from files whose names match a given pattern.
"""
flist = glob.glob(fpattern)
starttimes = []
endtimes = []
# first only read the header information
for fname in flist:
st = read(fname,headonly=True)
starttimes.append(st[0].stats.starttime.datetime)
endtimes.append(st[-1].stats.endtime.datetime)
# now read the stream from the files that contain the period
st = Stream()
for ind,fname in enumerate(flist):
if (starttimes[ind] < endtime) and (endtimes[ind] > starttime):
if trim:
st += read(fname,starttime=UTCDateTime(starttime),endtime=UTCDateTime(endtime))
else:
st += read(fname)
st.merge()
return st
def extractDataFromArchive(t1,
t2,
fileNames,
fileTimes,
wantedStaChas=[['*', '*']]):
# Return nothing if there is no data
if len(fileTimes) == 0:
return EmptyStream()
# Catch the case where the asked time range is completely outside the archive data availability
if t1 > fileTimes[-1, 1] or t2 < fileTimes[0, 0]:
return EmptyStream()
# Figure out what set of files are wanted
collectArgs = np.where((fileTimes[:, 0] <= t2)
& (fileTimes[:, 1] >= t1))[0]
stream = EmptyStream()
flagged = False
# Read in all of the information
for aFile in fileNames[collectArgs]:
# Flag to user if the archive structure has changed
if not os.path.exists(aFile):
flagged = True
continue
aStream = read(aFile)
for aSta, aCha in wantedStaChas:
stream += aStream.select(station=aSta, channel=aCha)
if flagged:
print('Archive structure as changed, reload the current archive')
# Merge traces which are adjacent
try:
stream.merge(method=1)
except:
stream = RemoveOddRateTraces(stream)
stream.merge(method=1)
# If any trace has masked values, split
if True in [isinstance(tr.data, np.ma.masked_array) for tr in stream]:
stream = stream.split()
# Trim to wanted times
stream.trim(UTCDateTime(t1), UTCDateTime(t2))
return stream
def getData(tstart, tend, opt):
"""
Download data from files in a folder, from IRIS, or a Earthworm waveserver
A note on SAC/miniSEED files: as this makes no assumptions about the naming scheme of
your data files, please ensure that your headers contain the correct SCNL information!
tstart: UTCDateTime of beginning of period of interest
tend: UTCDateTime of end of period of interest
opt: Options object describing station/run parameters
Returns ObsPy stream objects, one for cutting and the other for triggering
"""
nets = opt.network.split(',')
stas = opt.station.split(',')
locs = opt.location.split(',')
chas = opt.channel.split(',')
st = Stream()
if opt.server == 'file':
# Generate list of files
if opt.server == 'file':
flist = list(itertools.chain.from_iterable(glob.iglob(os.path.join(
root,opt.filepattern)) for root, dirs, files in os.walk(opt.searchdir)))
# Determine which subset of files to load based on start and end times and
# station name; we'll fully deal with stations below
flist_sub = []
for f in flist:
# Load header only
stmp = obspy.read(f, headonly=True)
# Check if station is contained in the stas list
if stmp[0].stats.station in stas:
# Check if contains either start or end time
ststart = stmp[0].stats.starttime
stend = stmp[-1].stats.endtime
if (ststart<=tstart and tstart<=stend) or (ststart<=tend and
tend<=stend) or (tstart<=stend and ststart<=tend):
flist_sub.append(f)
# Fully load data from file
stmp = Stream()
for f in flist_sub:
tmp = obspy.read(f, starttime=tstart, endtime=tend+opt.maxdt)
if len(tmp) > 0:
stmp = stmp.extend(tmp)
# Filter and merge
stmp = stmp.filter('bandpass', freqmin=opt.fmin, freqmax=opt.fmax, corners=2,
zerophase=True)
stmp = stmp.taper(0.05,type='hann',max_length=opt.mintrig)
for m in range(len(stmp)):
if stmp[m].stats.sampling_rate != opt.samprate:
stmp[m] = stmp[m].resample(opt.samprate)
stmp = stmp.merge(method=1, fill_value=0)
# Only grab stations/channels that we want and in order
netlist = []
stalist = []
chalist = []
loclist = []
for s in stmp:
stalist.append(s.stats.station)
chalist.append(s.stats.channel)
netlist.append(s.stats.network)
loclist.append(s.stats.location)
# Find match of SCNL in header or fill empty
for n in range(len(stas)):
for m in range(len(stalist)):
if (stas[n] in stalist[m] and chas[n] in chalist[m] and nets[n] in
netlist[m] and locs[n] in loclist[m]):
st = st.append(stmp[m])
if len(st) == n:
print("Couldn't find "+stas[n]+'.'+chas[n]+'.'+nets[n]+'.'+locs[n])
trtmp = Trace()
trtmp.stats.sampling_rate = opt.samprate
trtmp.stats.station = stas[n]
st = st.append(trtmp.copy())
else:
if '.' not in opt.server:
client = Client(opt.server)
else:
client = EWClient(opt.server, opt.port)
for n in range(len(stas)):
try:
stmp = client.get_waveforms(nets[n], stas[n], locs[n], chas[n],
tstart, tend+opt.maxdt)
for m in range(len(stmp)):
stmp[m].data = np.where(stmp[m].data == -2**31, 0, stmp[m].data) # replace -2**31 (Winston NaN token) w 0
stmp = stmp.filter('bandpass', freqmin=opt.fmin, freqmax=opt.fmax,
corners=2, zerophase=True)
stmp = stmp.taper(0.05,type='hann',max_length=opt.mintrig)
for m in range(len(stmp)):
if stmp[m].stats.sampling_rate != opt.samprate:
stmp[m] = stmp[m].resample(opt.samprate)
stmp = stmp.merge(method=1, fill_value=0)
except (obspy.clients.fdsn.header.FDSNException):
try: # try again
stmp = client.get_waveforms(nets[n], stas[n], locs[n], chas[n],
tstart, tend+opt.maxdt)
for m in range(len(stmp)):
stmp[m].data = np.where(stmp[m].data == -2**31, 0, stmp[m].data) # replace -2**31 (Winston NaN token) w 0
stmp = stmp.filter('bandpass', freqmin=opt.fmin, freqmax=opt.fmax,
corners=2, zerophase=True)
stmp = stmp.taper(0.05,type='hann',max_length=opt.mintrig)
for m in range(len(stmp)):
if stmp[m].stats.sampling_rate != opt.samprate:
stmp[m] = stmp[m].resample(opt.samprate)
stmp = stmp.merge(method=1, fill_value=0)
except (obspy.clients.fdsn.header.FDSNException):
print('No data found for {0}.{1}'.format(stas[n],nets[n]))
trtmp = Trace()
trtmp.stats.sampling_rate = opt.samprate
trtmp.stats.station = stas[n]
stmp = Stream().extend([trtmp.copy()])
# Last check for length; catches problem with empty waveserver
if len(stmp) != 1:
print('No data found for {0}.{1}'.format(stas[n],nets[n]))
trtmp = Trace()
trtmp.stats.sampling_rate = opt.samprate
trtmp.stats.station = stas[n]
stmp = Stream().extend([trtmp.copy()])
st.extend(stmp.copy())
# Edit 'start' time if using offset option
if opt.maxdt:
dts = np.fromstring(opt.offset, sep=',')
for n, tr in enumerate(st):
tr.stats.starttime = tr.stats.starttime-dts[n]
st = st.trim(starttime=tstart, endtime=tend, pad=True, fill_value=0)
stC = st.copy()
return st, stC
def grab_file_data(filepath, scnl, tstart, tend, fill_value=0):
import obspy
from obspy import Stream, Trace
import glob, os, itertools
stas = []
chas = []
nets = []
locs = []
for sta in scnl:
stas.append(sta.split('.')[0])
chas.append(sta.split('.')[1])
nets.append(sta.split('.')[2])
if len(sta) == 4:
locs.append(sta.split('.')[3])
else:
locs.append('')
st = Stream()
#if opt.server == 'file':
if True:
# Generate list of files
#if opt.server == 'file':
flist = list(
itertools.chain.from_iterable(
glob.iglob(os.path.join(root, "*"))
for root, dirs, files in os.walk(filepath)))
# "*" takes the place of wildcard lists, see REDPy documentation
# Determine which subset of files to load based on start and end times and
# station name; we'll fully deal with stations below
flist_sub = []
for f in flist:
# Load header only
stmp = obspy.read(f, headonly=True)
# Check if station is contained in the stas list
if stmp[0].stats.station in stas:
# Check if contains either start or end time
ststart = stmp[0].stats.starttime
stend = stmp[0].stats.endtime
if (ststart <= tstart and tstart <= stend) or (
ststart <= tend
and tend <= stend) or (tstart <= stend
and ststart <= tend):
flist_sub.append(f)
# Fully load data from file
stmp = Stream()
for f in flist_sub:
tmp = obspy.read(f, starttime=tstart, endtime=tend)
if len(tmp) > 0:
stmp = stmp.extend(tmp)
# Filter and merge
#stmp = stmp.filter('bandpass', freqmin=opt.fmin, freqmax=opt.fmax, corners=2,
# zerophase=True)
#stmp = stmp.taper(0.05,type='hann',max_length=opt.mintrig)
#for m in range(len(stmp)):
# if stmp[m].stats.sampling_rate != opt.samprate:
# stmp[m] = stmp[m].resample(opt.samprate)
stmp = stmp.merge(method=1, fill_value=fill_value)
# Only grab stations/channels that we want and in order
netlist = []
stalist = []
chalist = []
loclist = []
for s in stmp:
stalist.append(s.stats.station)
chalist.append(s.stats.channel)
netlist.append(s.stats.network)
loclist.append(s.stats.location)
# Find match of SCNL in header or fill empty
for n in range(len(stas)):
for m in range(len(stalist)):
if (stas[n] in stalist[m] and chas[n] in chalist[m]
and nets[n] in netlist[m] and locs[n] in loclist[m]):
st = st.append(stmp[m])
if len(st) == n:
print("Couldn't find " + stas[n] + '.' + chas[n] + '.' +
nets[n] + '.' + locs[n])
trtmp = Trace()
trtmp.stats.station = stas[n]
st = st.append(trtmp.copy())
if len(st) > 1:
if fill_value == 0 or fill_value == None:
st.detrend('demean')
st.taper(max_percentage=0.01)
st.merge(fill_value=fill_value)
st.trim(tstart, tend, pad=0)
st.detrend('demean')
print(st)
return st
def get_stream(datasource,
scnl,
tstart,
tend,
fill_value=0,
filepattern='*',
filter=None,
samprate=100,
verbose=False):
"""
Generalized (and more robust) way to retrieve waveform data through ObsPy
Download data from files in a folder, from IRIS, or a Earthworm waveserver
A note on SAC/miniSEED files: as this makes no assumptions about the naming scheme of
your data files, please ensure that your headers contain the correct SCNL information!
tstart: UTCDateTime of beginning of period of interest
tend: UTCDateTime of end of period of interest
filepattern='*'
You can specify a pattern for your files to reduce the files within the directory
searched. For example, filepattern=2019.06.*.mseed if your files are miniSEED files
named by date and you only want those from June 2019. Simple wildcarding is supported
(i.e., * and ?, [] for ranges of values or lists) but not full regular expressions.
samprate=100
Resamples all waveforms to the same sample rate.
Returns ObsPy stream objects
Based on code by Alicia Hotovec-Ellis and Aaron Wech.
Example:
>>> get_stream(['vdap.org', 16024], ['HSR.EHZ.CC.--'], '2004-09-28T00:00:00', '2004-09-28T01:00:00')
>>> get_stream(['file', '/Users/vdapseismo/data/'], ['HSR.EHZ.CC.--'], '2004-09-28T00:00:00', '2004-09-28T01:00:00')
>>> get_stream(['IRIS'], ['HSR.EHZ.CC.--'], '2004-09-28T00:00:00', '2004-09-28T01:00:00')
"""
from obspy import UTCDateTime
import obspy
from obspy.clients.fdsn import Client
from obspy.clients.earthworm import Client as EWClient
from obspy.core.trace import Trace
from obspy.core.stream import Stream
from obspy.signal.trigger import coincidence_trigger
import numpy as np
from scipy import stats
from scipy.fftpack import fft
import glob, os, itertools
#print(datasource)
#print(scnl)
#print(tstart)
#print(tend)
tstart = UTCDateTime(tstart)
tend = UTCDateTime(tend)
nets = []
stas = []
locs = []
chas = []
for s in scnl:
#print(s)
nets.append(s.split('.')[2])
stas.append(s.split('.')[0])
locs.append(s.split('.')[3])
chas.append(s.split('.')[1])
st = Stream()
if '/' in datasource:
# Retrieve data from file structure
flist = list(
itertools.chain.from_iterable(
glob.iglob(os.path.join(root, filepattern))
for root, dirs, files in os.walk(datasource)))
# Determine which subset of files to load based on start and end times and
# station name; we'll fully deal with stations below
flist_sub = []
for f in flist:
# Load header only
stmp = obspy.read(f, headonly=True)
# Check if station is contained in the stas list
if stmp[0].stats.station in stas:
# Check if contains either start or end time
ststart = stmp[0].stats.starttime
stend = stmp[0].stats.endtime
if (ststart <= tstart and tstart <= stend) or (
ststart <= tend
and tend <= stend) or (tstart <= stend
and ststart <= tend):
flist_sub.append(f)
# Fully load data from file
stmp = Stream()
for f in flist_sub:
tmp = obspy.read(f, starttime=tstart, endtime=tend)
if len(tmp) > 0:
stmp = stmp.extend(tmp)
# merge
stmp = stmp.taper(max_percentage=0.01)
for m in range(len(stmp)):
if stmp[m].stats.sampling_rate != samprate:
stmp[m] = stmp[m].resample(samprate)
stmp = stmp.merge(method=1, fill_value=fill_value)
# Only grab stations/channels that we want and in order
netlist = []
stalist = []
chalist = []
loclist = []
for s in stmp:
stalist.append(s.stats.station)
chalist.append(s.stats.channel)
netlist.append(s.stats.network)
loclist.append(s.stats.location)
# Find match of SCNL in header or fill empty
for n in range(len(stas)):
for m in range(len(stalist)):
if (stas[n] in stalist[m] and chas[n] in chalist[m]
and nets[n] in netlist[m] and locs[n] in loclist[m]):
st = st.append(stmp[m])
if len(st) == n:
print('No data found for {}.{}.{}.{}'.format(
stas[n], chas[n], nets[n], locs[n]))
trtmp = Trace()
trtmp.stats.sampling_rate = samprate
trtmp.stats.station = stas[n]
st = st.append(trtmp.copy())
else:
# retrieve data from server
if '.' not in datasource:
client = Client(datasource)
else:
datasource = datasource.split(':')
client = EWClient(datasource[0], int(datasource[1]))
for n in range(len(stas)):
try:
stmp = client.get_waveforms(nets[n], stas[n], locs[n], chas[n],
tstart, tend)
for m in range(len(stmp)):
#stmp[m].data = np.ma.masked_where(stmp[m].data == -2**31, stmp[m].data) # masks out all values of -2**31 (Winston NaN Token)
#stmp[m] = stmp[m].split().merge(method=0, fill_value='interpolate')[0] # splits trace at masked values; then re-merges using linear interpolation
stmp[m].data = np.where(stmp[m].data == -2**31, 0,
stmp[m].data)
if stmp[m].stats.sampling_rate != samprate:
stmp[m] = stmp[m].resample(samprate)
stmp = stmp.taper(max_percentage=0.01)
stmp = stmp.merge(method=1, fill_value=fill_value)
except (obspy.clients.fdsn.header.FDSNException):
try: # try again
stmp = client.get_waveforms(nets[n], stas[n], locs[n],
chas[n], tstart, tend)
for m in range(len(stmp)):
#stmp[m].data = np.ma.masked_where(stmp[m].data == -2**31, stmp[m].data) # masks out all values of -2**31 (Winston NaN Token)
#stmp[m] = stmp[m].split().merge(method=0, fill_value='interpolate')[0] # splits trace at masked values; then re-merges using linear interpolation
stmp[m].data = np.where(stmp[m].data == -2**31, 0,
stmp[m].data)
if stmp[m].stats.sampling_rate != samprate:
stmp[m] = stmp[m].resample(samprate)
stmp = stmp.taper(max_percentage=0.01)
stmp = stmp.merge(method=1, fill_value=fill_value)
except (obspy.clients.fdsn.header.FDSNException):
print('No data found for {0}.{1}'.format(stas[n], nets[n]))
trtmp = Trace()
trtmp.stats.sampling_rate = samprate
trtmp.stats.station = stas[n]
stmp = Stream().extend([trtmp.copy()])
# Last check for length; catches problem with empty waveserver
if len(stmp) != 1:
print('No data found for {}.{}.{}.{}'.format(
stas[n], chas[n], nets[n], locs[n]))
trtmp = Trace()
trtmp.stats.sampling_rate = samprate
trtmp.stats.station = stas[n]
stmp = Stream().extend([trtmp.copy()])
st.extend(stmp.copy())
st = st.trim(starttime=tstart,
endtime=tend,
pad=True,
fill_value=fill_value)
return st
# get list of all days and of unique days
days_all = np.array([
datetime.strptime(a[-8:], '%Y.%j').strftime('%Y.%m.%d') for a in dayfiles
])
days = np.unique(days_all)
for dy in days:
file_list = dayfiles[np.where(
days_all == dy)] # files for this day, all stations
st = Stream()
for fl in file_list:
if fl not in bad_files and os.stat(fl).st_size != 0:
st += read(fl)
if len(st) > 0:
# decimate or resample to 1Hz
if np.all([tr.stats.sampling_rate == 100 for tr in st]):
st.decimate(100, no_filter=True)
else:
st.resample(1.0, no_filter=True)
st.merge(method=1, fill_value=0) # make sure 1 trace per sta/comp
for tr in st: # rename channels to L* to match [edited] dataless
tr.stats.channel = 'L' + tr.stats.channel[1:]
# write miniseed
ofile = os.path.join(mseed_dir_out,'EN-%s.%s.%s.mseed' % \
(dy.split('.')[0],dy.split('.')[1],dy.split('.')[2]))
st.write(ofile, format='MSEED')
def ProcessStream(stream):
n_day = 0
for trace in stream:
if trace.stats.channel != channel:
continue
net = trace.stats.network
sta = trace.stats.station
station_path = sacPath + sta + '-' + net
if not os.path.exists(station_path):
os.makedirs(station_path)
starttime = trace.stats.starttime
endtime = trace.stats.endtime
day_gap = datetime.date(
endtime.year, endtime.month, endtime.day) - datetime.date(
starttime.year, starttime.month, starttime.day)
time_day = [starttime]
for ii in range(day_gap.days):
splittime = datetime.date(time_day[ii].year, time_day[ii].month,
time_day[ii].day) + datetime.timedelta(1)
splittime = str(splittime.year) + '-' + str(
splittime.month) + '-' + str(splittime.day) + 'T00:00:00'
splittime = UTCDateTime(splittime)
time_day.append(splittime)
time_day.append(endtime)
for ii in range(len(time_day) - 1):
starttime = time_day[ii]
endtime = time_day[ii + 1]
st_day = trace.slice(starttime, endtime, nearest_sample=False)
if st_day.stats.npts < 10:
continue
day_name = sacPath + sta + '-' + net + '/' + sta + '-' + net + '-' + str(
starttime.year) + '-' + str(
starttime.julday).zfill(3) + '-' + channel + '.SAC'
if os.path.exists(day_name):
st_tmp = read(day_name, format='SAC')
if int(st_tmp[0].stats.sampling_rate) != int(resamplingRate):
resampling_rate = int(
round(st_tmp[0].stats.sampling_rate / resamplingRate))
st_tmp[0].decimate(resampling_rate)
day_name_tmp = day_name + '.tmp'
st_day.write(day_name_tmp, format='SAC')
st_day = read(day_name_tmp, format='SAC')
os.remove(day_name_tmp)
if int(st_day[0].stats.sampling_rate) != int(resamplingRate):
resampling_rate = int(
round(st_day[0].stats.sampling_rate / resamplingRate))
if resampling_rate == 100:
st_day[0].decimate(10, no_filter=True)
st_day[0].decimate(10, no_filter=True)
else:
st_day[0].decimate(resampling_rate, no_filter=True)
if np.abs(st_day[0].stats.sampling_rate -
resamplingRate) > 0.01:
st_day[0].resample(resamplingRate)
else:
st_day[0].stats.sampling_rate = resamplingRate
st_tmp[0].data = st_tmp[0].data.astype(np.float32)
st_day[0].data = st_day[0].data.astype(np.float32)
merge_data = Stream()
merge_data.append(st_tmp[0])
merge_data.append(st_day[0])
merge_data.sort(['starttime'])
merge_data.merge(method=1, fill_value='latest')
os.remove(day_name)
merge_data.write(day_name, format='SAC')
else:
if int(st_day.stats.sampling_rate) != int(resamplingRate):
resampling_rate = int(
round(st_day.stats.sampling_rate / resamplingRate))
if resampling_rate == 100:
st_day.decimate(10, no_filter=True)
st_day.decimate(10, no_filter=True)
else:
st_day.decimate(resampling_rate, no_filter=True)
if np.abs(st_day.stats.sampling_rate - resamplingRate) > 0.01:
st_day.resample(resamplingRate)
else:
st_day.stats.sampling_rate = resamplingRate
st_day.write(day_name, format='SAC')
n_day = n_day + 1
print(sta + ' now julday is: ' + str(starttime.date), n_day)
def getData(tstart, tend, opt):
"""
Download data from files in a folder, from IRIS, or a Earthworm waveserver
A note on SAC/miniSEED files: as this makes no assumptions about the naming scheme of
your data files, please ensure that your headers contain the correct SCNL information!
tstart: UTCDateTime of beginning of period of interest
tend: UTCDateTime of end of period of interest
opt: Options object describing station/run parameters
Returns ObsPy stream objects, one for cutting and the other for triggering
"""
nets = opt.network.split(',')
stas = opt.station.split(',')
locs = opt.location.split(',')
chas = opt.channel.split(',')
st = Stream()
if opt.server == 'SAC' or opt.server == 'miniSEED':
# Generate list of files
if opt.server == 'SAC':
flist = glob.glob(opt.sacdir + '*.sac') + glob.glob(opt.sacdir +
'*.SAC')
elif opt.server == 'miniSEED':
flist = glob.glob(opt.mseeddir +
'*.mseed') + glob.glob(opt.mseeddir + '*.MSEED')
# Load data from file
stmp = Stream()
for f in flist:
tmp = obspy.read(f, starttime=tstart, endtime=tend)
if len(tmp) > 0:
stmp = stmp.extend(tmp)
# Filter and merge
stmp = stmp.filter('bandpass',
freqmin=opt.fmin,
freqmax=opt.fmax,
corners=2,
zerophase=True)
stmp = stmp.taper(0.05, type='hann', max_length=opt.mintrig)
for m in range(len(stmp)):
if stmp[m].stats.sampling_rate != opt.samprate:
stmp[m] = stmp[m].resample(opt.samprate)
stmp = stmp.merge(method=1, fill_value=0)
# Only grab stations/channels that we want and in order
netlist = []
stalist = []
chalist = []
loclist = []
for s in stmp:
stalist.append(s.stats.station)
chalist.append(s.stats.channel)
netlist.append(s.stats.network)
loclist.append(s.stats.location)
# Find match of SCNL in header or fill empty
for n in range(len(stas)):
for m in range(len(stalist)):
if (stas[n] in stalist[m] and chas[n] in chalist[m]
and nets[n] in netlist[m] and locs[n] in loclist[m]):
st = st.append(stmp[m])
if len(st) == n:
print("Couldn't find " + stas[n] + '.' + chas[n] + '.' +
nets[n] + '.' + locs[n])
trtmp = Trace()
trtmp.stats.sampling_rate = opt.samprate
trtmp.stats.station = stas[n]
st = st.append(trtmp.copy())
else:
if '.' not in opt.server:
client = Client(opt.server)
else:
client = EWClient(opt.server, opt.port)
for n in range(len(stas)):
try:
stmp = client.get_waveforms(nets[n], stas[n], locs[n], chas[n],
tstart, tend)
stmp = stmp.filter('bandpass',
freqmin=opt.fmin,
freqmax=opt.fmax,
corners=2,
zerophase=True)
stmp = stmp.taper(0.05, type='hann', max_length=opt.mintrig)
for m in range(len(stmp)):
if stmp[m].stats.sampling_rate != opt.samprate:
stmp[m] = stmp[m].resample(opt.samprate)
stmp = stmp.merge(method=1, fill_value=0)
except (obspy.fdsn.header.FDSNException):
try: # try again
stmp = client.get_waveforms(nets[n], stas[n], locs[n],
chas[n], tstart, tend)
stmp = stmp.filter('bandpass',
freqmin=opt.fmin,
freqmax=opt.fmax,
corners=2,
zerophase=True)
stmp = stmp.taper(0.05,
type='hann',
max_length=opt.mintrig)
for m in range(len(stmp)):
if stmp[m].stats.sampling_rate != opt.samprate:
stmp[m] = stmp[m].resample(opt.samprate)
stmp = stmp.merge(method=1, fill_value=0)
except (obspy.fdsn.header.FDSNException):
print('No data found for {0}.{1}'.format(stas[n], nets[n]))
trtmp = Trace()
trtmp.stats.sampling_rate = opt.samprate
trtmp.stats.station = stas[n]
stmp = Stream().extend([trtmp.copy()])
st.extend(stmp.copy())
st = st.trim(starttime=tstart, endtime=tend, pad=True, fill_value=0)
stC = st.copy()
return st, stC
def getData(tstart, tend, opt):
"""
Download data from files in a folder, from IRIS, or a Earthworm waveserver
A note on SAC/miniSEED files: as this makes no assumptions about the naming scheme of
your data files, please ensure that your headers contain the correct SCNL information!
tstart: UTCDateTime of beginning of period of interest
tend: UTCDateTime of end of period of interest
opt: Options object describing station/run parameters
Returns ObsPy stream objects, one for cutting and the other for triggering
"""
nets = opt.network.split(',')
stas = opt.station.split(',')
locs = opt.location.split(',')
chas = opt.channel.split(',')
st = Stream()
if opt.server == 'SAC' or opt.server == 'miniSEED':
# Generate list of files
if opt.server == 'SAC':
flist = list(itertools.chain.from_iterable(glob.iglob(os.path.join(
root,'*.sac')) for root, dirs, files in os.walk(opt.sacdir)))+list(
itertools.chain.from_iterable(glob.iglob(os.path.join(
root,'*.SAC')) for root, dirs, files in os.walk(opt.sacdir)))
elif opt.server == 'miniSEED':
flist = list(itertools.chain.from_iterable(glob.iglob(os.path.join(
root,'*.mseed')) for root, dirs, files in os.walk(opt.mseeddir)))+list(
itertools.chain.from_iterable(glob.iglob(os.path.join(
root,'*.MSEED')) for root, dirs, files in os.walk(opt.mseeddir)))
# Determine which subset of files to load based on start and end times and
# station name; we'll fully deal with stations below
flist_sub = []
for f in flist:
# Load header only
stmp = obspy.read(f, headonly=True)
# Check if station is contained in the stas list
if stmp[0].stats.station in stas:
# Check if contains either start or end time
ststart = stmp[0].stats.starttime
stend = stmp[0].stats.endtime
if (ststart<=tstart and tstart<=stend) or (ststart<=tend and
tend<=stend) or (tstart<=stend and ststart<=tend):
flist_sub.append(f)
# Fully load data from file
stmp = Stream()
for f in flist_sub:
tmp = obspy.read(f, starttime=tstart, endtime=tend+opt.maxdt)
if len(tmp) > 0:
stmp = stmp.extend(tmp)
# Filter and merge
stmp = stmp.filter('bandpass', freqmin=opt.fmin, freqmax=opt.fmax, corners=2,
zerophase=True)
stmp = stmp.taper(0.05,type='hann',max_length=opt.mintrig)
for m in range(len(stmp)):
if stmp[m].stats.sampling_rate != opt.samprate:
stmp[m] = stmp[m].resample(opt.samprate)
stmp = stmp.merge(method=1, fill_value=0)
# Only grab stations/channels that we want and in order
netlist = []
stalist = []
chalist = []
loclist = []
for s in stmp:
stalist.append(s.stats.station)
chalist.append(s.stats.channel)
netlist.append(s.stats.network)
loclist.append(s.stats.location)
# Find match of SCNL in header or fill empty
for n in range(len(stas)):
for m in range(len(stalist)):
if (stas[n] in stalist[m] and chas[n] in chalist[m] and nets[n] in
netlist[m] and locs[n] in loclist[m]):
st = st.append(stmp[m])
if len(st) == n:
print("Couldn't find "+stas[n]+'.'+chas[n]+'.'+nets[n]+'.'+locs[n])
trtmp = Trace()
trtmp.stats.sampling_rate = opt.samprate
trtmp.stats.station = stas[n]
st = st.append(trtmp.copy())
else:
if '.' not in opt.server:
client = Client(opt.server)
else:
client = EWClient(opt.server, opt.port)
for n in range(len(stas)):
try:
stmp = client.get_waveforms(nets[n], stas[n], locs[n], chas[n],
tstart, tend+opt.maxdt)
stmp = stmp.filter('bandpass', freqmin=opt.fmin, freqmax=opt.fmax,
corners=2, zerophase=True)
stmp = stmp.taper(0.05,type='hann',max_length=opt.mintrig)
for m in range(len(stmp)):
if stmp[m].stats.sampling_rate != opt.samprate:
stmp[m] = stmp[m].resample(opt.samprate)
stmp = stmp.merge(method=1, fill_value=0)
except (obspy.clients.fdsn.header.FDSNException):
try: # try again
stmp = client.get_waveforms(nets[n], stas[n], locs[n], chas[n],
tstart, tend+opt.maxdt)
stmp = stmp.filter('bandpass', freqmin=opt.fmin, freqmax=opt.fmax,
corners=2, zerophase=True)
stmp = stmp.taper(0.05,type='hann',max_length=opt.mintrig)
for m in range(len(stmp)):
if stmp[m].stats.sampling_rate != opt.samprate:
stmp[m] = stmp[m].resample(opt.samprate)
stmp = stmp.merge(method=1, fill_value=0)
except (obspy.clients.fdsn.header.FDSNException):
print('No data found for {0}.{1}'.format(stas[n],nets[n]))
trtmp = Trace()
trtmp.stats.sampling_rate = opt.samprate
trtmp.stats.station = stas[n]
stmp = Stream().extend([trtmp.copy()])
# Last check for length; catches problem with empty waveserver
if len(stmp) != 1:
print('No data found for {0}.{1}'.format(stas[n],nets[n]))
trtmp = Trace()
trtmp.stats.sampling_rate = opt.samprate
trtmp.stats.station = stas[n]
stmp = Stream().extend([trtmp.copy()])
st.extend(stmp.copy())
# Edit 'start' time if using offset option
if opt.maxdt:
dts = np.fromstring(opt.offset, sep=',')
for n, tr in enumerate(st):
tr.stats.starttime = tr.stats.starttime-dts[n]
st = st.trim(starttime=tstart, endtime=tend, pad=True, fill_value=0)
stC = st.copy()
return st, stC
print(file)
site_idx = sites.index(file.split('/')[-1].split('.')[0])
channel_idx = channels.index(file.split('/')[-1].split('.')[2][-3:])
split_files[site_idx][channel_idx].append(file)
# Load in data for each site and perform noise analysis
for m, site in enumerate(sites):
print('Parsing data for site: ' + site)
for n, channel in enumerate(channels):
streams = Stream()
for o in range(len(split_files[m][n])):
print('Parsing file:')
print(split_files[m][n][o])
streams += obspy.read(split_files[m][n][o])
print('Merging streams...')
streams.merge()
print('Current data is:')
print(streams)
# Build probabilistic power spectral density objects for each trace
all_ppsds = []
all_ppsd_names = []
for stream in streams:
print('Calculating PPSDs for stream:')
print(stream)
ppsds = []
ppsd_names = []
metadata = Parser(metadata_directory + stream.stats.station +
stream.stats.channel[-1:] + '.seed')
ppsd = PPSD(stream.stats, metadata)
ppsd.add(stream)
if (stations_ok) and (tr.stats.station not in stations):
continue
if (locations_ok) and (tr.stats.location not in locations):
continue
if (channels_ok) and (tr.stats.channel not in channels):
continue
st.append(tr)
if st.count() > 0:
# If the starttime is given then it trims the resulting traces
if starttime:
st.trim(starttime=starttime,
endtime=endtime,
nearest_sample=nearest_sample)
st.merge(method=1, fill_value=0, interpolation_samples=1)
else:
print "Empty stream"
n_trace = st.count()
return st, n_trace
if BC_UI:
class _stream_dbread_view(HasTraits):
time_interval = Float(30.0)
networks = List(Str, value=['PF', 'YA'])
stations = List(Str, value=['FOR', 'UV05'])
locations = List(Str, value=['00', '10'])
continue
if (stations_ok) and (tr.stats.station not in stations):
continue
if (locations_ok) and (tr.stats.location not in locations):
continue
if (channels_ok) and (tr.stats.channel not in channels):
continue
st.append(tr)
if st.count() > 0:
# If the starttime is given then it trims the resulting traces
if starttime:
st.trim(starttime=starttime, endtime=endtime,
nearest_sample=nearest_sample)
st.merge(method=1, fill_value=0, interpolation_samples=1)
else:
print "Empty stream"
n_trace = st.count()
return st, n_trace
if BC_UI:
class _stream_dbread_view(HasTraits):
time_interval = Float(30.0)
networks = List(Str, value=['PF', 'YA'])
stations = List(Str, value=['FOR', 'UV05'])
locations = List(Str, value=['00', '10'])
channels = List(Str, value=['HHZ', 'HLZ'])
