def picker(st):
"""docstring for picker"""
s_picks=[np.nan]*2
df = st[0].stats.sampling_rate
npts = st[0].stats.npts
for i,cmp in enumerate(['L','Q','T']):
tmp=st.select(component=cmp)
trace=tmp[0]
# calculate cft of zDetect for 20 sample window
cft=zDetect(trace,20)
#mincft=min(cft)
maxcft=max(cft)
#meancft=mean(cft)
#print mincft,maxcft,(maxcft+mincft)/2,meancft
if cmp=='L': # P-trace: care less about this one, go ahead and pick
picks=triggerOnset(cft,0,0)
# plotTrigger(trace, cft, 0, 0)
p_pick=np.nan
for pick,end in picks:
if npts-pick<100:
break
else:
stcut = st.copy()
# cut trace from pick to 100 samples past pick
startcut=stcut[0].stats.starttime+dt.timedelta(seconds= pick/df)
endcut=startcut + dt.timedelta(seconds=100/df)
stcut = stcut.slice(startcut,endcut)
stcut.sort()
azi,inc,lin,plan=flinn(stcut)
#print "inc, lin =", inc, lin
if inc<40 and lin>0.6:
# good enough for a p-pick
p_pick=pick
break
else: # S-traces: only pick if the signal is relatively large
s_picks[i-1]=np.nan
if maxcft>6:
picks=triggerOnset(cft,4,1)
for pick,end in picks:
if npts-pick<100:
break
else:
stcut =st.copy()
# cut trace from pick to 100 samples past pick
startcut=stcut[0].stats.starttime+dt.timedelta(seconds= pick/df)
endcut=startcut + dt.timedelta(seconds=100/df)
stcut = stcut.slice(startcut,endcut)
stcut.sort()
azi,inc,lin,plan=flinn(stcut)
if inc>50 and lin>0.75:
# good enough for an s-pick
s_picks[i-1]=pick
break
s_pick=np.nanmin(s_picks)
return (p_pick-50)/df,[(i-50)/df for i in s_picks]
from obspy.core import read
from obspy.signal.trigger import zDetect, plotTrigger
trace = read("http://examples.obspy.org/ev0_6.a01.gse2")[0]
df = trace.stats.sampling_rate
cft = zDetect(trace.data, int(10. * df))
plotTrigger(trace, cft, -0.4, -0.3)