#!/usr/bin/env python3

# -*- coding: utf-8 -*-

"""

Created on Mon Jan 20 16:34:12 2020

@author: root

"""

import os

from collections import OrderedDict

import numpy as np

import obspy

import scipy.signal as sgn

import matplotlib.pyplot as plt

import matplotlib.mlab as mlab

from scipy import integrate

from obspy.clients.earthworm import Client

from obspy import UTCDateTime

from obspy.signal.trigger import classic_sta_lta, recursive_sta_lta

from obspy.signal.trigger import plot_trigger, trigger_onset

from obspy import Stream

from numpy import argmax

# STATION, CHANNEL (DDF --> 400 Hz), NETWWORK AND LOCATION CODES

sta = 'FG8' # STATION

cha = 'BHZ' # CHANNEL

net = 'GI' #

loc = '00' # location, it depends mostly of which network you are in.

client = Client('138.253.112.23', 16022) # ip, port - ip's 138.253.113.19 or 138.253.112.23

t1 = UTCDateTime(2018, 5, 20, 2, 58, 20) #the format is year:day_of_the_year:month

t2 = t1 + 90

st_ref2 = Stream()

st_ref2 = client.get_waveforms(net, sta, loc, cha, t1 , t2)

st_ref2.detrend(type='linear')

st_ref2.detrend(type='demean')

st_ref2.filter(type='bandpass',freqmin=0.5, freqmax=5)

st_ref2.plot(color='r',starttime=t1, endtime=t2)

ref2_st = st_ref2[0].data

#%%

# STATION, CHANNEL (DDF --> 400 Hz), NETWWORK AND LOCATION CODES

sta = 'FG8' # STATION

cha = 'BDF' # CHANNEL

net = 'GI' #

loc = '02' # location, it depends mostly of which network you are in.

#client = Client('138.253.112.23', 16022) # ip, port - ip's 138.253.113.19 or 138.253.112.23

#%%

t1 = UTCDateTime(2019, 10, 1, 0, 0, 0) #the format is year:day_of_the_year:month

t2 = t1 + 24*60*60

st_inf = Stream()

st_inf = client.get_waveforms(net, sta, loc, cha, t1 , t2)

st_inf.detrend(type='linear')

st_inf.detrend(type='demean')

print('inf in')

st_inf1 = st_inf.copy()

fb = st_inf1#.filter(type='bandpass',freqmin=0.1, freqmax=22)

trace=fb[0]

cha2 = 'BHZ' # CHANNEL

loc2 = '00'

sts = Stream()

sts = client.get_waveforms(net, sta, loc2, cha2, t1 , t2)

sts.detrend(type='linear')

sts.detrend(type='demean')

print('seis in')

sts1 = sts.copy()

sts2 = sts.copy()

sts0 = sts.copy()

fb1 = sts0.filter(type='bandpass',freqmin=0.5, freqmax=22)

fb2 = sts1.filter(type='bandpass',freqmin=12, freqmax=22)

fb3 = sts2.filter(type='bandpass',freqmin=0.5, freqmax=5)

trace1=fb1[0]

trace2=fb2[0]

trace3=fb3[0]

#%%

#window endpoints

start= t1 + 10 #time window start

end= t2

trs = trace.slice(starttime = start , endtime= end)

trs.plot(type='relative',color='b')#, starttime=start , endtime=end)

sr = trace.stats.sampling_rate

nsta=int(0.5*sr)

nlta=int(25*sr)

stream=trs.data

cft=recursive_sta_lta(stream, nsta, nlta)

trig_on=20

trig_off=1

plot_trigger(trs, cft, trig_on, trig_off)

on_off = trigger_onset(cft,trig_on,trig_off)

#%%

catalogue = np.zeros(shape=(0,1))

num=0

shift2 = 1000

for x in range(0,len(on_off)):

event = trs[on_off[x,0]-500:on_off[x,0]+1000]

eventsh = trace2[on_off[x,0]-2500:on_off[x,0]+4500]

eventsl = trace3[on_off[x,0]-2500:on_off[x,0]+4500]

eventsa = trace1[on_off[x,0]-2500:on_off[x,0]+4500]

start = t1 + on_off[x,0]/sr -20

end = start + 40

trace_fft = trace.slice(starttime = start , endtime= end)

window=end-start

tr_data=trace_fft.data

m=np.mean(tr_data)

tr_data = tr_data-m

famp = abs(np.fft.fft(tr_data))

# double peak ratio

fps=int(len(famp)/(sr/0.5))

midp=int(len(famp)/(sr/2))

spe=int(len(famp)/(sr/4.5))

f_peak_m = np.mean(famp[fps:midp])/(window*sr)

s_peak_m = np.mean(famp[midp:spe])/(window*sr)

peak_r = f_peak_m/s_peak_m

# 2Hz drop ratio

fpbs=int(len(famp)/(sr/0.5))

fpbe=int(len(famp)/(sr/1.8))

spbs=int(len(famp)/(sr/2.2))

spbe=int(len(famp)/(sr/3.5))

peaks_ma = np.mean(famp[fpbs:fpbe])/(window*sr)

peaks_mb = np.mean(famp[spbs:spbe])/(window*sr)

peaks_m = (peaks_ma + peaks_mb)/2

trough_m = np.mean(famp[fpbe:spbs])/(window*sr)

trough_r = peaks_m/trough_m

top_vl,top3,corell3 = corel(abs(ref2_st),abs(eventsl),shift2)

# dom,cf, bw = freq_info_NqV(event_a,start,end,sr)

# top_v,top,corell = corel(event_a,event_l,shift)

# top_vh,top2,corell2 = corel(abs(ref_st),abs(event_h),shift2)

# top_vl,top3,corell3 = corel(abs(ref2_st),abs(event_l),shift2)

# plt.figure()

# plt.plot(event,color='k')

# plt.title(trough_r)

#

# plt.figure()

# plt.plot(eventsl,color='b')

# plt.plot(eventsh,color='r')

# plt.title(UTCDateTime(start + on_off[x,0]/sr ))

#

# plt.figure()

# plt.plot(event_l,color='k')

# plt.title(top_vl)

#

# plt.figure()

# plt.plot(event_h,color='r')

# plt.title(top_vh)

# if trough_r < 1:

# plt.figure()

# plt.plot(event,color='k')

# plt.title(trough_r)

if top_vl > 0.3 and trough_r > 1:

catalogue = np.lib.pad(catalogue, ((0,1),(0,0)), 'constant', constant_values=(0))

catalogue[num]= start + on_off[x,0]/sr

num+=1

plt.figure()

plt.plot(event,color='b')

plt.title(UTCDateTime(start + on_off[x,0]/sr ))

# plt.figure()

# plt.plot(eventsl,color='b')

# plt.plot(eventsh,color='r')

#

# plt.title(UTCDateTime(start + on_off[x,0]/sr ))

# else:

# plt.figure()

# plt.plot(event,color='k')

# plt.title(trough_r)

# plt.figure()

# plt.plot(eventsa,color='r')

# plt.plot(eventsh,color='k')

# plt.title(top_vl)

#%%

print('number of detections =',num)

print('number of rejections =',len(on_off)-num)

for x in range(0,len(catalogue)):

print(UTCDateTime(catalogue[x]))