#!/usr/bin/env python3

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

"""

Created on Tue Mar 26 18:05:51 2019

@author: william

"""

import obspy

from obspy import read

import numpy as np

from numpy import genfromtxt

import matplotlib.pyplot as plt

from obspy import Stream

from obspy import UTCDateTime

from obspy.clients.earthworm import Client

#%%

secondary= genfromtxt("//Users/william/Documents/scanner/analysis/secondaries_list.csv", delimiter=',',skip_header=1)

corr_list= genfromtxt("/Users/william/Documents/scanner/analysis/LB01_inf_correltion_list_6s_v2.csv", delimiter=',',skip_header=1)

event_times= genfromtxt("/Users/william/Documents/scanner/analysis/LB01_inf_correltion_event_details_6s_v2.csv", delimiter=',',skip_header=1)

corr_m= genfromtxt("/Users/william/Documents/scanner/analysis/LB01_inf_correltion_matrix_6s_v2.csv", delimiter=',',skip_header=1)

#%%

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

sta = 'LB01' # STATION

cha = 'HDF' # CHANNEL

net = 'Z4' #

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

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

for x in range(0,20):

t1 = UTCDateTime(secondary[x][1]) #the format is year:day_of_the_year:month

t2 = t1 + 6

st = Stream()

st = client.get_waveforms(net, sta, '', cha, t1-20 , t2+60)

#print(st)

st.detrend(type='linear')

st.detrend(type='demean')

st.filter(type='bandpass',freqmin=0.2, freqmax=2)

st.plot(color='b',starttime=t1-15, endtime=t2+30)

#%%

count = 0

#for x in range(0,100):

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

primary = secondary[x][0]

second = secondary[x][1]

nearp,indp=find_nearest(event_times[:], primary)

nears,inds=find_nearest(event_times[:], second)

if abs(primary - nearp) < 60:

if abs(second-nears) < 60:

print(nearp)

count += 1

else:

print("no secondary in list")

else:

print("no primary in list")

print(len(secondary))

print(count)

#%%

num =0

for y in range(1,len(event_times)):

if event_times[y] - event_times[y-1] < 600:

num += 1

print(num)

#%%

last_p = event_times[0]

prim_i = 0

numb_s=0

numb_a=0

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

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

cot = 0

for x in range(1,len(event_times)):

#for x in range(1,250):

print('x=',x)

for y in range(0,x):

cot += 1

if x - y < 3:

if 0 < event_times[x]-event_times[y] < 600:

if event_times[x]-event_times[y-1] > 600:

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

ps_corr[numb_s] = corr_m[x][y]

numb_s +=1

else:

if event_times[x]-event_times[y-2] > 600:

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

ps_corr[numb_s] = corr_m[x][y-1]

numb_s +=1

else:

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

a_corr[numb_a] = corr_m[x][y]

numb_a +=1

else:

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

a_corr[numb_a] = corr_m[x][y]

numb_a +=1

#%%

s_mean = np.mean(ps_corr)

s_var = np.var(ps_corr)

s_n = len(ps_corr)

a_mean = np.mean(a_corr)

a_var = np.var(a_corr)

a_n = len(a_corr)

t_calc = (s_mean - a_mean)/np.sqrt((s_var/s_n)+(a_var/a_n))

print("t_calc (any pair to secondary) = ", t_calc) # = 6.800 (3dp) - significant to 0.1%

s_std = np.std(ps_corr)

a_std = np.std(a_corr)

print('secondary mean = ', s_mean)

print('normal mean = ', a_mean)

print('secondary variance = ', s_var)

print('normal variance = ', a_var)

print('secondary standard deviation = ', s_std)

print('normal standard deviation = ', a_std)

#%%

np.savetxt("/Users/william/Documents/scanner/analysis/primary_secondary_correlation_v2.csv", ps_corr,delimiter=",",header="")

np.savetxt("/Users/william/Documents/scanner/analysis/all_explosions_correlation_v2.csv", a_corr,delimiter=",",header="")

#%%

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

numb_n =0

for x in range(1,len(event_times)):

if event_times[x]-event_times[x-1] > 600:

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

n_corr[numb_n] = corr_m[x][x-1]

numb_n +=1

n_mean = np.mean(n_corr)

n_var = np.var(n_corr)

n_n = len(n_corr)

t_calc = (s_mean - n_mean)/np.sqrt((s_var/s_n)+(n_var/n_n))

print("t_calc (next to secondary) = ", t_calc) # = 4.067 (3dp) - significant to 0.1%

#%%

t_calc = (a_mean - n_mean)/np.sqrt((a_var/a_n)+(n_var/n_n))

print("t_calc (next to any) = ", abs(t_calc)) # = 6.215 (3dp) - significant to 0.1%