-
Notifications
You must be signed in to change notification settings - Fork 3
/
disaggregation_oq_sources.py
203 lines (183 loc) · 6.52 KB
/
disaggregation_oq_sources.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
#Given a stochastic set of ruptures as exported from OQ and a (mean_disagg.csv) disaggregation result
#(processed using the /home/mhaas/RIESGOS/disaggregation/createPlot.py routine)
#returns the stochastic set of events associated with the poe of the disaggregation bin it belongs to
import os
import pandas
import numpy as np
#import scipy
#import time
#t0=time.time()
##OQruptures
#rup = pandas.read_csv("ruptures_3411.csv",skiprows=1,delimiter='\t')
##disaggregation
#dr = pandas.read_csv("mean_disagg_0.1.csv")
#
#def match_row(data,ref):
# '''
# given data (x,y,z) matches it to reference (x,y,z)
# and returns index of row in ref matching row in data
# columns have to be in same order!
# NOTE: CHANGED RETURNS POEs
# '''
# cd = data.columns
# cr = ref.columns
# poes=[]
# #idxs=[]
# for i in range(len(data)):
# row = data.iloc[i]
# #match
# try:
# #idxs.append(ref[(ref[cr[0]]==row[cd[0]])&(ref[cr[1]]==row[cd[1]])&(ref[cr[2]]==row[cd[2]])].index[0])
# poes.append(float(ref[(ref[cr[0]]==row[cd[0]])&(ref[cr[1]]==row[cd[1]])&(ref[cr[2]]==row[cd[2]])].poe))
# except:
# #print(row)
# #pass
# poes.append(0.)
# #return idxs
# return poes
#
def oqrup2cat(ruptures,dtype='deaggregation',provider='GFZ'):
'''
Converts a set of OQ ruptures to a catalog
'''
#initialize
index = [i for i in range(len(ruptures))]
columns=['eventID', 'Agency', 'Identifier', 'year', 'month', 'day', 'hour', 'minute', 'second', 'timeError', 'longitude', 'latitude','SemiMajor90', 'SemiMinor90', 'ErrorStrike', 'depth', 'depthError', 'magnitude', 'sigmaMagnitude','rake','dip','strike','type', 'probability', 'fuzzy']
catalog=pandas.DataFrame(index=index,columns=columns)
#add values
catalog.eventID = ruptures.rupid
catalog.Agency = provider
catalog.longitude = ruptures.centroid_lon
catalog.latitude = ruptures.centroid_lat
catalog.depth = ruptures.centroid_depth
catalog.magnitude = ruptures.mag
catalog.type = dtype
#not necessarily defined
try:
catalog.strike = ruptures.strike
except:
pass
try:
catalog.dip = ruptures.dip
except:
pass
try:
catalog.rake = ruptures.rake
except:
pass
try:
catalog.probability = ruptures.poe
except:
pass
return catalog
def binning_xyz(data,px,py,pz):
'''
given pandas data frame (data x,y,z) and bins dy,dy,dz
returns binned data
'''
xyz=data.copy()
cols=xyz.columns
#rounds to bin precision
xyz[cols[0]] = xyz[cols[0]]/px
xyz[cols[1]] = xyz[cols[1]]/py
xyz[cols[2]] = xyz[cols[2]]/pz
xyz = xyz.round()
xyz[cols[0]] = xyz[cols[0]]*px
xyz[cols[1]] = xyz[cols[1]]*py
xyz[cols[2]] = xyz[cols[2]]*pz
return xyz
#FIXME: Add uncertainty here, i.e., calculate sigmas from all matching events/or just use half bins
def return_random_event(events,disagg,seed=42):
'''
Per unique bin returns index of single random event and poe for corresponding disaggregation bin (which have assigned bins)
'''
idxs = []
poe = []
#go through bins of disaggregation
for i in range(len(disagg)):
seed+=i
row = disagg.iloc[i]
#get events
matches = events[(abs(events.longitude-row.Lon)<10**-5)&(abs(events.latitude-row.Lat)<10**-5)&(abs(events.magnitude-row.Mag)<10**-5)]
#append single random sampled idx
n=len(matches)
if n>0:
np.random.seed(seed)
idx=np.random.randint(0,n,1)[0]
idxs.append(matches.iloc[idx].name)
poe.append(row.poe)
return [idxs,poe]
def match_disaggregation(ruptures,lat,lon,poe):
'''
Given a set of ruptures, a target with longitude/latitude,
and a target exceedance probability (e.g., 0.1 = 10%) for 50 years return period
picks up corresponding deaggregation and selects a single random event
from the rupture for each bin
'''
#read deaggregation sites
filepath=os.path.dirname(__file__)
sites_filename = os.path.join(filepath,"sites.csv")
sites = pandas.read_csv(sites_filename)
#find closest match to target
slon= [sites.iloc[i].lon for i,v in enumerate(sites.lon) if abs(v-lon)==min(abs(sites.lon - lon))][0]
slon= [sites.iloc[i].lon for i,v in enumerate(sites.lon) if abs(v-lon)==min(abs(sites.lon - lon))][0]
slat= [sites.iloc[i].lat for i,v in enumerate(sites.lat) if abs(v-lat)==min(abs(sites.lat - lat))][0]
sid = int(sites[(sites.lon==slon) & (sites.lat==slat)].sid)
#get deaggregation
disagg_filename = os.path.join(filepath,"mean_disagg.csv")
dr = pandas.read_csv(disagg_filename)
#get that for specified hazard level and site
dr = dr[(dr.sid==sid) & (dr.poe50y==poe)]
#determine precision
plon = round(min(np.diff(dr.Lon.unique())),5)
plat = round(min(np.diff(dr.Lat.unique())),5)
pmag = round(min(np.diff(dr.Mag.unique())),5)
#bin the ruptures
bins = binning_xyz(ruptures[['longitude','latitude','magnitude']],plon,plat,pmag)
#take only those with non-zero poe
dr = dr[dr.poe>0]
#select events
idxs,poe = return_random_event(bins,dr,seed=42)
matches = ruptures.loc[idxs]
matches['probability']=poe
return matches
#make sure no index problems for following conversion
#matches.to_csv('matches.csv',index=False)
#matches = matches.reset_index()
##convert to catalog style
#catalog = oqrup2cat(matches,provider='GFZ')
##save matches
#catalog.to_csv('catalog.csv',index=False)
#print(time.time()-t0)
##determine precision of disaggregation (up to 5 digits)
#plon = round(min(np.diff(dr.Lon.unique())),5)
#plat = round(min(np.diff(dr.Lat.unique())),5)
#pmag = round(min(np.diff(dr.Mag.unique())),5)
#
##bin the ruptures
#bins = binning_xyz(rup[['centroid_lon','centroid_lat','mag']],plon,plat,pmag)
#associate each event in OQruptures with poe in dr
#rup['poe'] = 0.
#get matches
#poes,idxs = match_row(bins,dr)
#rup['poe'] = match_row(bins,dr)
##TAKE ONLY NON-ZERO
#print('WARNING: CONSIDERING ONLY DISAGGREGATION BINS WITH POE > 0')
#dr = dr[dr.poe>0]
#
##selects events
#idxs,poe = return_random_event(bins,dr,seed=42)
#matches = rup.loc[idxs]
#matches['poe']=poe
##make sure no index problems for following conversion
#matches.to_csv('matches.csv',index=False)
#matches = matches.reset_index()
#
##convert to catalog style
#catalog = oqrup2cat(matches,provider='GFZ')
#
#
##save matches
#catalog.to_csv('catalog.csv',index=False)
#
#print(time.time()-t0)