# -*- coding: utf-8 -*-
# vim: tabstop=4 shiftwidth=4 softtabstop=4
#
# Copyright (C) 2014-2018 GEM Foundation
#
# OpenQuake is free software: you can redistribute it and/or modify it
# under the terms of the GNU Affero General Public License as published
# by the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
#
# OpenQuake is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU Affero General Public License for more details.
#
# You should have received a copy of the GNU Affero General Public License
# along with OpenQuake. If not, see <http://www.gnu.org/licenses/>.
import os
import csv
import copy
import zlib
import shutil
import random
import zipfile
import logging
import tempfile
import operator
import functools
import configparser
import collections
import numpy
from openquake.baselib import performance, hdf5, parallel
from openquake.baselib.general import (
AccumDict, DictArray, deprecated, random_filter)
from openquake.baselib.python3compat import decode, zip
from openquake.baselib.node import Node
from openquake.hazardlib.const import StdDev
from openquake.hazardlib.calc.filters import (
split_sources, SourceFilter, RtreeFilter)
from openquake.hazardlib.calc.gmf import CorrelationButNoInterIntraStdDevs
from openquake.hazardlib import (
geo, site, imt, valid, sourceconverter, nrml, InvalidFile)
from openquake.hazardlib.geo.mesh import point3d
from openquake.hazardlib.probability_map import ProbabilityMap
from openquake.risklib import asset, riskinput
from openquake.risklib.riskmodels import get_risk_models
from openquake.commonlib.oqvalidation import OqParam
from openquake.commonlib import logictree, source, writers
# the following is quite arbitrary, it gives output weights that I like (MS)
NORMALIZATION_FACTOR = 1E-2
RUPTURES_PER_BLOCK = 10000 # used in split_filter
TWO16 = 2 ** 16 # 65,536
F32 = numpy.float32
F64 = numpy.float64
U16 = numpy.uint16
U32 = numpy.uint32
U64 = numpy.uint64
Site = collections.namedtuple('Site', 'sid lon lat')
[docs]class DuplicatedPoint(Exception):
"""
Raised when reading a CSV file with duplicated (lon, lat) pairs
"""
[docs]def collect_files(dirpath, cond=lambda fullname: True):
"""
Recursively collect the files contained inside dirpath.
:param dirpath: path to a readable directory
:param cond: condition on the path to collect the file
"""
files = []
for fname in os.listdir(dirpath):
fullname = os.path.join(dirpath, fname)
if os.path.isdir(fullname): # navigate inside
files.extend(collect_files(fullname))
else: # collect files
if cond(fullname):
files.append(fullname)
return files
[docs]def normalize(key, fnames, base_path):
input_type, _ext = key.rsplit('_', 1)
filenames = []
for val in fnames:
if os.path.isabs(val):
raise ValueError('%s=%s is an absolute path' % (key, val))
val = os.path.normpath(os.path.join(base_path, val))
if (key in ('source_model_logic_tree_file', 'exposure_file') and
not os.path.exists(val)):
zpath = val[:-4] + '.zip'
if not os.path.exists(zpath):
raise OSError('No such file: %s or %s' % (val, zpath))
with zipfile.ZipFile(zpath) as archive:
archive.extractall(os.path.dirname(zpath))
filenames.append(val)
return input_type, filenames
def _update(params, items, base_path):
for key, value in items:
if key in ('hazard_curves_csv', 'site_model_file', 'exposure_file'):
input_type, fnames = normalize(key, value.split(), base_path)
params['inputs'][input_type] = fnames
elif key.endswith(('_file', '_csv', '_hdf5')):
if value:
input_type, [fname] = normalize(key, [value], base_path)
params['inputs'][input_type] = fname
elif isinstance(value, str) and value.endswith('.hdf5'):
# for the reqv feature
fname = os.path.normpath(os.path.join(base_path, value))
try:
reqv = params['inputs']['reqv']
except KeyError:
params['inputs']['reqv'] = {key: fname}
else:
reqv.update({key: fname})
else:
params[key] = value
[docs]def get_params(job_inis, **kw):
"""
Parse one or more INI-style config files.
:param job_inis:
List of configuration files (or list containing a single zip archive)
:param kw:
Optionally override some parameters
:returns:
A dictionary of parameters
"""
input_zip = None
if len(job_inis) == 1 and job_inis[0].endswith('.zip'):
input_zip = job_inis[0]
job_inis = extract_from_zip(
job_inis[0], ['job_hazard.ini', 'job_haz.ini',
'job.ini', 'job_risk.ini'])
not_found = [ini for ini in job_inis if not os.path.exists(ini)]
if not_found: # something was not found
raise IOError('File not found: %s' % not_found[0])
cp = configparser.ConfigParser()
cp.read(job_inis)
# directory containing the config files we're parsing
job_ini = os.path.abspath(job_inis[0])
base_path = decode(os.path.dirname(job_ini))
params = dict(base_path=base_path, inputs={'job_ini': job_ini})
if input_zip:
params['inputs']['input_zip'] = os.path.abspath(input_zip)
for sect in cp.sections():
_update(params, cp.items(sect), base_path)
_update(params, kw.items(), base_path) # override on demand
if params['inputs'].get('reqv'):
# using pointsource_distance=0 because of the reqv approximation
params['pointsource_distance'] = '0'
return params
[docs]def get_oqparam(job_ini, pkg=None, calculators=None, hc_id=None, validate=1,
**kw):
"""
Parse a dictionary of parameters from an INI-style config file.
:param job_ini:
Path to configuration file/archive or dictionary of parameters
:param pkg:
Python package where to find the configuration file (optional)
:param calculators:
Sequence of calculator names (optional) used to restrict the
valid choices for `calculation_mode`
:param hc_id:
Not None only when called from a post calculation
:param validate:
Flag. By default it is true and the parameters are validated
:param kw:
String-valued keyword arguments used to override the job.ini parameters
:returns:
An :class:`openquake.commonlib.oqvalidation.OqParam` instance
containing the validate and casted parameters/values parsed from
the job.ini file as well as a subdictionary 'inputs' containing
absolute paths to all of the files referenced in the job.ini, keyed by
the parameter name.
"""
# UGLY: this is here to avoid circular imports
from openquake.calculators import base
OqParam.calculation_mode.validator.choices = tuple(
calculators or base.calculators)
if not isinstance(job_ini, dict):
basedir = os.path.dirname(pkg.__file__) if pkg else ''
job_ini = get_params([os.path.join(basedir, job_ini)])
if hc_id:
job_ini.update(hazard_calculation_id=str(hc_id))
job_ini.update(kw)
oqparam = OqParam(**job_ini)
if validate:
oqparam.validate()
return oqparam
pmap = None # set as side effect when the user reads hazard_curves from a file
# the hazard curves format does not split the site locations from the data (an
# unhappy legacy design choice that I fixed in the GMFs CSV format only) thus
# this hack is necessary, otherwise we would have to parse the file twice
exposure = None # set as side effect when the user reads the site mesh
# this hack is necessary, otherwise we would have to parse the exposure twice
gmfs, eids = None, None # set as a sided effect when reading gmfs.xml
# this hack is necessary, otherwise we would have to parse the file twice
[docs]def read_csv(fname, sep=','):
"""
:param fname: a CSV file with an header and float fields
:param sep: separato (default the comma)
:return: a structured array of floats
"""
with open(fname, encoding='utf-8-sig') as f:
header = next(f).strip().split(sep)
dt = numpy.dtype([(h, float) for h in header])
return numpy.loadtxt(f, dt, delimiter=sep)
[docs]def get_mesh(oqparam):
"""
Extract the mesh of points to compute from the sites,
the sites_csv, or the region.
:param oqparam:
an :class:`openquake.commonlib.oqvalidation.OqParam` instance
"""
global pmap, exposure, gmfs, eids
if 'exposure' in oqparam.inputs and exposure is None:
# read it only once
exposure = get_exposure(oqparam)
if oqparam.sites:
return geo.Mesh.from_coords(oqparam.sites)
elif 'sites' in oqparam.inputs:
fname = oqparam.inputs['sites']
header = get_csv_header(fname)
if 'lon' in header:
data = []
for i, row in enumerate(
csv.DictReader(open(fname, 'U', encoding='utf-8-sig'))):
if header[0] == 'site_id' and row['site_id'] != str(i):
raise InvalidFile('%s: expected site_id=%d, got %s' % (
fname, i, row['site_id']))
data.append(' '.join([row['lon'], row['lat']]))
elif 'gmfs' in oqparam.inputs:
raise InvalidFile('Missing header in %(sites)s' % oqparam.inputs)
else:
data = [line.replace(',', ' ')
for line in open(fname, 'U', encoding='utf-8-sig')]
coords = valid.coordinates(','.join(data))
start, stop = oqparam.sites_slice
c = (coords[start:stop] if header[0] == 'site_id'
else sorted(coords[start:stop]))
return geo.Mesh.from_coords(c)
elif 'hazard_curves' in oqparam.inputs:
fname = oqparam.inputs['hazard_curves']
if isinstance(fname, list): # for csv
mesh, pmap = get_pmap_from_csv(oqparam, fname)
elif fname.endswith('.xml'):
mesh, pmap = get_pmap_from_nrml(oqparam, fname)
else:
raise NotImplementedError('Reading from %s' % fname)
return mesh
elif 'gmfs' in oqparam.inputs:
eids, gmfs = _get_gmfs(oqparam) # sets oqparam.sites
return geo.Mesh.from_coords(oqparam.sites)
elif oqparam.region_grid_spacing:
if oqparam.region:
poly = geo.Polygon.from_wkt(oqparam.region)
elif 'site_model' in oqparam.inputs:
sm = get_site_model(oqparam)
poly = geo.Mesh(sm['lon'], sm['lat']).get_convex_hull()
elif exposure:
poly = exposure.mesh.get_convex_hull()
else:
raise InvalidFile('There is a grid spacing but not a region, '
'nor a site model, nor an exposure in %s' %
oqparam.inputs['job_ini'])
try:
mesh = poly.dilate(oqparam.region_grid_spacing).discretize(
oqparam.region_grid_spacing)
return geo.Mesh.from_coords(zip(mesh.lons, mesh.lats))
except Exception:
raise ValueError(
'Could not discretize region with grid spacing '
'%(region_grid_spacing)s' % vars(oqparam))
elif 'exposure' in oqparam.inputs:
return exposure.mesh
[docs]def get_site_model(oqparam):
"""
Convert the NRML file into an array of site parameters.
:param oqparam:
an :class:`openquake.commonlib.oqvalidation.OqParam` instance
:returns:
an array with fields lon, lat, vs30, ...
"""
req_site_params = get_gsim_lt(oqparam).req_site_params
arrays = []
for fname in oqparam.inputs['site_model']:
if isinstance(fname, str) and fname.endswith('.csv'):
sm = read_csv(fname)
if 'site_id' in sm.dtype.names:
raise InvalidFile('%s: you passed a sites.csv file instead of '
'a site_model.csv file!' % fname)
arrays.append(sm)
continue
nodes = nrml.read(fname).siteModel
params = [valid.site_param(node.attrib) for node in nodes]
missing = req_site_params - set(params[0])
if 'vs30measured' in missing: # use a default of False
missing -= {'vs30measured'}
for param in params:
param['vs30measured'] = False
if 'backarc' in missing: # use a default of False
missing -= {'backarc'}
for param in params:
param['backarc'] = False
if missing:
raise InvalidFile('%s: missing parameter %s' %
(oqparam.inputs['site_model'],
', '.join(missing)))
# NB: the sorted in sorted(params[0]) is essential, otherwise there is
# an heisenbug in scenario/test_case_4
site_model_dt = numpy.dtype([(p, site.site_param_dt[p])
for p in sorted(params[0])])
sm = numpy.array([tuple(param[name] for name in site_model_dt.names)
for param in params], site_model_dt)
arrays.append(sm)
return numpy.concatenate(arrays)
[docs]def get_site_collection(oqparam):
"""
Returns a SiteCollection instance by looking at the points and the
site model defined by the configuration parameters.
:param oqparam:
an :class:`openquake.commonlib.oqvalidation.OqParam` instance
"""
mesh = get_mesh(oqparam)
req_site_params = get_gsim_lt(oqparam).req_site_params
if oqparam.inputs.get('site_model'):
sm = get_site_model(oqparam)
try:
# in the future we could have elevation in the site model
depth = sm['depth']
except ValueError:
# this is the normal case
depth = None
if mesh is None:
# extract the site collection directly from the site model
sitecol = site.SiteCollection.from_points(
sm['lon'], sm['lat'], depth, sm, req_site_params)
else:
sitecol = site.SiteCollection.from_points(
mesh.lons, mesh.lats, mesh.depths, None, req_site_params)
if oqparam.region_grid_spacing:
logging.info('Reducing the grid sites to the site '
'parameters within the grid spacing')
sitecol, params, _ = geo.utils.assoc(
sm, sitecol, oqparam.region_grid_spacing * 1.414, 'filter')
sitecol.make_complete()
else:
# associate the site parameters to the sites without
# discarding any site but warning for far away parameters
sc, params, _ = geo.utils.assoc(
sm, sitecol, oqparam.max_site_model_distance, 'warn')
for name in req_site_params:
if name in ('vs30measured', 'backarc') \
and name not in params.dtype.names:
sitecol._set(name, 0) # the default
else:
sitecol._set(name, params[name])
elif mesh is None and oqparam.ground_motion_fields:
raise InvalidFile('You are missing sites.csv or site_model.csv in %s'
% oqparam.inputs['job_ini'])
elif mesh is None:
# a None sitecol is okay when computing the ruptures only
return
else: # use the default site params
sitecol = site.SiteCollection.from_points(
mesh.lons, mesh.lats, mesh.depths, oqparam, req_site_params)
ss = os.environ.get('OQ_SAMPLE_SITES')
if ss:
# debugging tip to reduce the size of a calculation
# OQ_SAMPLE_SITES=.1 oq engine --run job.ini
# will run a computation with 10 times less sites
sitecol.array = numpy.array(random_filter(sitecol.array, float(ss)))
sitecol.make_complete()
return sitecol
[docs]def get_gsim_lt(oqparam, trts=['*']):
"""
:param oqparam:
an :class:`openquake.commonlib.oqvalidation.OqParam` instance
:param trts:
a sequence of tectonic region types as strings; trts=['*']
means that there is no filtering
:returns:
a GsimLogicTree instance obtained by filtering on the provided
tectonic region types.
"""
if 'gsim_logic_tree' not in oqparam.inputs:
return logictree.GsimLogicTree.from_(oqparam.gsim)
gsim_file = os.path.join(
oqparam.base_path, oqparam.inputs['gsim_logic_tree'])
gsim_lt = logictree.GsimLogicTree(gsim_file, trts)
gmfcorr = oqparam.correl_model
for trt, gsims in gsim_lt.values.items():
for gsim in gsims:
if gmfcorr and (gsim.DEFINED_FOR_STANDARD_DEVIATION_TYPES ==
{StdDev.TOTAL}):
raise CorrelationButNoInterIntraStdDevs(gmfcorr, gsim)
trts = set(oqparam.minimum_magnitude) - {'default'}
expected_trts = set(gsim_lt.values)
assert trts <= expected_trts, (trts, expected_trts)
return gsim_lt
[docs]def get_gsims(oqparam):
"""
Return an ordered list of GSIM instances from the gsim name in the
configuration file or from the gsim logic tree file.
:param oqparam:
an :class:`openquake.commonlib.oqvalidation.OqParam` instance
"""
return [rlz.value[0] for rlz in get_gsim_lt(oqparam)]
[docs]def get_rlzs_by_gsim(oqparam):
"""
Return an ordered dictionary gsim -> [realization index]. Work for
gsim logic trees with a single tectonic region type.
"""
cinfo = source.CompositionInfo.fake(get_gsim_lt(oqparam))
ra = cinfo.get_rlzs_assoc()
dic = {}
for rlzi, gsim_by_trt in enumerate(ra.gsim_by_trt):
dic[gsim_by_trt['*']] = [rlzi]
return dic
[docs]def get_rupture(oqparam):
"""
Read the `rupture_model` file and by filter the site collection
:param oqparam:
an :class:`openquake.commonlib.oqvalidation.OqParam` instance
:returns:
an hazardlib rupture
"""
rup_model = oqparam.inputs['rupture_model']
[rup_node] = nrml.read(rup_model)
conv = sourceconverter.RuptureConverter(
oqparam.rupture_mesh_spacing, oqparam.complex_fault_mesh_spacing)
rup = conv.convert_node(rup_node)
rup.tectonic_region_type = '*' # there is not TRT for scenario ruptures
rup.serial = oqparam.random_seed
return rup
[docs]def get_source_model_lt(oqparam):
"""
:param oqparam:
an :class:`openquake.commonlib.oqvalidation.OqParam` instance
:returns:
a :class:`openquake.commonlib.logictree.SourceModelLogicTree`
instance
"""
fname = oqparam.inputs.get('source_model_logic_tree')
if fname:
# NB: converting the random_seed into an integer is needed on Windows
return logictree.SourceModelLogicTree(
fname, validate=False, seed=int(oqparam.random_seed),
num_samples=oqparam.number_of_logic_tree_samples)
return logictree.FakeSmlt(oqparam.inputs['source_model'],
int(oqparam.random_seed),
oqparam.number_of_logic_tree_samples)
[docs]def check_nonparametric_sources(fname, smodel, investigation_time):
"""
:param fname:
full path to a source model file
:param smodel:
source model object
:param investigation_time:
investigation_time to compare with in the case of
nonparametric sources
:returns:
the nonparametric sources in the model
:raises:
a ValueError if the investigation_time is different from the expected
"""
# NonParametricSeismicSources
np = [src for sg in smodel.src_groups for src in sg
if hasattr(src, 'data')]
if np and smodel.investigation_time != investigation_time:
raise ValueError(
'The source model %s contains an investigation_time '
'of %s, while the job.ini has %s' % (
fname, smodel.investigation_time, investigation_time))
return np
[docs]class SourceModelFactory(object):
def __init__(self):
self.fname_hits = collections.Counter() # fname -> number of calls
self.changes = 0
def __call__(self, fname, sm, apply_uncertainties, investigation_time):
"""
:param fname:
the full pathname of a source model file
:param sm:
the original source model
:param apply_uncertainties:
a function modifying the sources (or None)
:param investigation_time:
the investigation_time in the job.ini
:returns:
a copy of the original source model with changed sources, if any
"""
check_nonparametric_sources(fname, sm, investigation_time)
newsm = nrml.SourceModel(
[], sm.name, sm.investigation_time, sm.start_time)
for group in sm:
newgroup = apply_uncertainties(group)
newsm.src_groups.append(newgroup)
if hasattr(newgroup, 'changed') and newgroup.changed.any():
self.changes += newgroup.changed.sum()
for src, changed in zip(newgroup, newgroup.changed):
# redoing count_ruptures can be slow
if changed:
src.num_ruptures = src.count_ruptures()
self.fname_hits[fname] += 1
return newsm
source_info_dt = numpy.dtype([
('grp_id', numpy.uint16), # 0
('source_id', hdf5.vstr), # 1
('code', (numpy.string_, 1)), # 2
('gidx1', numpy.uint32), # 3
('gidx2', numpy.uint32), # 4
('num_ruptures', numpy.uint32), # 5
('calc_time', numpy.float32), # 6
('split_time', numpy.float32), # 7
('num_sites', numpy.float32), # 8
('num_split', numpy.uint32), # 9
('weight', numpy.float32), # 10
])
[docs]def store_sm(smodel, hdf5path, monitor):
"""
:param smodel: a :class:`openquake.hazardlib.nrml.SourceModel` instance
:param hdf5path: path to an hdf5 file (cache_XXX.hdf5)
:param monitor: a Monitor instance with an .hdf5 attribute
"""
h5 = monitor.hdf5
with monitor('store source model'):
sources = h5['source_info']
source_geom = h5['source_geom']
gid = 0
for sg in smodel:
if hdf5path:
with hdf5.File(hdf5path, 'r+') as hdf5cache:
hdf5cache['grp-%02d' % sg.id] = sg
srcs = []
geoms = []
for src in sg:
srcgeom = src.geom()
n = len(srcgeom)
geom = numpy.zeros(n, point3d)
geom['lon'], geom['lat'], geom['depth'] = srcgeom.T
srcs.append((sg.id, src.source_id, src.code, gid, gid + n,
src.num_ruptures, 0, 0, 0, 0, 0))
geoms.append(geom)
gid += n
if geoms:
hdf5.extend(source_geom, numpy.concatenate(geoms))
if sources:
hdf5.extend(sources, numpy.array(srcs, source_info_dt))
[docs]def get_source_models(oqparam, gsim_lt, source_model_lt, monitor,
in_memory=True, srcfilter=None):
"""
Build all the source models generated by the logic tree.
:param oqparam:
an :class:`openquake.commonlib.oqvalidation.OqParam` instance
:param gsim_lt:
a :class:`openquake.commonlib.logictree.GsimLogicTree` instance
:param source_model_lt:
a :class:`openquake.commonlib.logictree.SourceModelLogicTree` instance
:param monitor:
a `openquake.baselib.performance.Monitor` instance
:param in_memory:
if True, keep in memory the sources, else just collect the TRTs
:param srcfilter:
a SourceFilter instance with an .hdf5path pointing to the cache file
:returns:
an iterator over :class:`openquake.commonlib.logictree.LtSourceModel`
tuples
"""
make_sm = SourceModelFactory()
spinning_off = oqparam.pointsource_distance == {'default': 0.0}
if spinning_off:
logging.info('Removing nodal plane and hypocenter distributions')
dist = 'no' if os.environ.get('OQ_DISTRIBUTE') == 'no' else 'processpool'
smlt_dir = os.path.dirname(source_model_lt.filename)
converter = sourceconverter.SourceConverter(
oqparam.investigation_time,
oqparam.rupture_mesh_spacing,
oqparam.complex_fault_mesh_spacing,
oqparam.width_of_mfd_bin,
oqparam.area_source_discretization,
oqparam.minimum_magnitude,
not spinning_off,
oqparam.source_id)
if oqparam.calculation_mode.startswith('ucerf'):
[grp] = nrml.to_python(oqparam.inputs["source_model"], converter)
elif in_memory:
logging.info('Reading the source model(s) in parallel')
smap = parallel.Starmap(
nrml.read_source_models, monitor=monitor, distribute=dist)
for sm in source_model_lt.gen_source_models(gsim_lt):
for name in sm.names.split():
fname = os.path.abspath(os.path.join(smlt_dir, name))
smap.submit([fname], converter)
dic = {sm.fname: sm for sm in smap}
# consider only the effective realizations
nr = 0
idx = 0
grp_id = 0
if monitor.hdf5:
sources = hdf5.create(monitor.hdf5, 'source_info', source_info_dt)
hdf5.create(monitor.hdf5, 'source_geom', point3d)
hdf5path = (getattr(srcfilter, 'hdf5path', None)
if oqparam.prefilter_sources == 'no' else None)
source_ids = set()
for sm in source_model_lt.gen_source_models(gsim_lt):
apply_unc = functools.partial(
source_model_lt.apply_uncertainties, sm.path)
src_groups = []
for name in sm.names.split():
fname = os.path.abspath(os.path.join(smlt_dir, name))
if oqparam.calculation_mode.startswith('ucerf'):
sg = copy.copy(grp)
sg.id = grp_id
src = sg[0].new(sm.ordinal, sm.names) # one source
source_ids.add(src.source_id)
src.src_group_id = grp_id
src.id = idx
if oqparam.number_of_logic_tree_samples:
src.samples = sm.samples
sg.sources = [src]
src_groups.append(sg)
idx += 1
grp_id += 1
data = [((sg.id, src.source_id, src.code, 0, 0,
src.num_ruptures, 0, 0, 0, 0, 0))]
hdf5.extend(sources, numpy.array(data, source_info_dt))
elif in_memory:
newsm = make_sm(fname, dic[fname], apply_unc,
oqparam.investigation_time)
for sg in newsm:
nr += sum(src.num_ruptures for src in sg)
# sample a source for each group
if os.environ.get('OQ_SAMPLE_SOURCES'):
sg.sources = random_filtered_sources(
sg.sources, srcfilter, sg.id + oqparam.random_seed)
for src in sg:
source_ids.add(src.source_id)
src.src_group_id = grp_id
src.id = idx
idx += 1
sg.id = grp_id
grp_id += 1
src_groups.append(sg)
if monitor.hdf5:
store_sm(newsm, hdf5path, monitor)
else: # just collect the TRT models
groups = logictree.read_source_groups(fname)
for group in groups:
source_ids.update(src['id'] for src in group)
src_groups.extend(groups)
if grp_id >= TWO16:
# the limit is really needed only for event based calculations
raise ValueError('There is a limit of %d src groups!' % TWO16)
for srcid in source_model_lt.info.applytosources:
if srcid not in source_ids:
raise ValueError(
'The source %s is not in the source model, please fix '
'applyToSources in %s or the source model' %
(srcid, source_model_lt.filename))
num_sources = sum(len(sg.sources) for sg in src_groups)
sm.src_groups = src_groups
trts = [mod.trt for mod in src_groups]
source_model_lt.tectonic_region_types.update(trts)
logging.info(
'Processed source model %d with %d gsim path(s) and %d '
'sources', sm.ordinal + 1, sm.num_gsim_paths, num_sources)
gsim_file = oqparam.inputs.get('gsim_logic_tree')
if gsim_file: # check TRTs
for src_group in src_groups:
if src_group.trt not in gsim_lt.values:
raise ValueError(
"Found in %r a tectonic region type %r inconsistent "
"with the ones in %r" % (sm, src_group.trt, gsim_file))
yield sm
logging.info('The composite source model has {:,d} ruptures'.format(nr))
# log if some source file is being used more than once
dupl = 0
for fname, hits in make_sm.fname_hits.items():
if hits > 1:
logging.info('%s has been considered %d times', fname, hits)
if not make_sm.changes:
dupl += hits
if (dupl and not oqparam.optimize_same_id_sources and
'event_based' not in oqparam.calculation_mode):
logging.warn('You are doing redundant calculations: please make sure '
'that different sources have different IDs and set '
'optimize_same_id_sources=true in your .ini file')
if make_sm.changes:
logging.info('Applied %d changes to the composite source model',
make_sm.changes)
[docs]def getid(src):
try:
return src.source_id
except AttributeError:
return src['id']
[docs]def random_filtered_sources(sources, srcfilter, seed):
"""
:param sources: a list of sources
:param srcfilte: a SourceFilter instance
:param seed: a random seed
:returns: an empty list or a list with a single filtered source
"""
random.seed(seed)
while sources:
src = random.choice(sources)
if srcfilter.get_close_sites(src) is not None:
return [src]
sources.remove(src)
return []
[docs]def get_composite_source_model(oqparam, monitor=None, in_memory=True,
srcfilter=SourceFilter(None, {})):
"""
Parse the XML and build a complete composite source model in memory.
:param oqparam:
an :class:`openquake.commonlib.oqvalidation.OqParam` instance
:param monitor:
a `openquake.baselib.performance.Monitor` instance
:param in_memory:
if False, just parse the XML without instantiating the sources
:param srcfilter:
if not None, use it to prefilter the sources
"""
source_model_lt = get_source_model_lt(oqparam)
trts = source_model_lt.get_trts()
trts_lower = {trt.lower() for trt in trts}
reqv = oqparam.inputs.get('reqv', {})
for trt in reqv: # these are lowercase because they come from the job.ini
if trt not in trts_lower:
raise ValueError('Unknown TRT=%s in %s [reqv]' %
(trt, oqparam.inputs['job_ini']))
gsim_lt = get_gsim_lt(oqparam, trts or ['*'])
p = source_model_lt.num_paths * gsim_lt.get_num_paths()
if oqparam.number_of_logic_tree_samples:
logging.info('Considering {:,d} logic tree paths out of {:,d}'.format(
oqparam.number_of_logic_tree_samples, p))
else: # full enumeration
if ('event_based' in oqparam.calculation_mode
and p > oqparam.max_potential_paths):
raise ValueError(
'There are too many potential logic tree paths (%d) '
'use sampling instead of full enumeration' % p)
logging.info('Potential number of logic tree paths = {:,d}'.format(p))
if source_model_lt.on_each_source:
logging.info('There is a logic tree on each source')
if monitor is None:
monitor = performance.Monitor()
smodels = []
for source_model in get_source_models(
oqparam, gsim_lt, source_model_lt, monitor, in_memory, srcfilter):
for src_group in source_model.src_groups:
src_group.sources = sorted(src_group, key=getid)
for src in src_group:
# there are two cases depending on the flag in_memory:
# 1) src is a hazardlib source and has a src_group_id
# attribute; in that case the source has to be numbered
# 2) src is a Node object, then nothing must be done
if isinstance(src, Node):
continue
smodels.append(source_model)
csm = source.CompositeSourceModel(gsim_lt, source_model_lt, smodels,
oqparam.optimize_same_id_sources)
for sm in csm.source_models:
counter = collections.Counter()
for sg in sm.src_groups:
for srcid in map(getid, sg):
counter[srcid] += 1
dupl = [srcid for srcid in counter if counter[srcid] > 1]
if dupl:
raise nrml.DuplicatedID('Found duplicated source IDs in %s: %s'
% (sm, dupl))
if not in_memory:
return csm
if 'event_based' in oqparam.calculation_mode:
# initialize the rupture serial numbers before splitting/filtering; in
# this way the serials are independent from the site collection
csm.init_serials(oqparam.ses_seed)
if oqparam.disagg_by_src:
csm = csm.grp_by_src() # one group per source
csm.info.gsim_lt.check_imts(oqparam.imtls)
if monitor.hdf5:
csm.info.gsim_lt.store_gmpe_tables(monitor.hdf5)
if (oqparam.prefilter_sources == 'rtree' and
oqparam.calculation_mode != 'ucerf_classical'):
# rtree can be used only with processpool, otherwise one gets an
# RTreeError: Error in "Index_Create": Spatial Index Error:
# IllegalArgumentException: SpatialIndex::DiskStorageManager:
# Index/Data file cannot be read/writen.
srcfilter = RtreeFilter(srcfilter.sitecol,
oqparam.maximum_distance,
srcfilter.hdf5path)
if (srcfilter and oqparam.prefilter_sources != 'no' and
oqparam.calculation_mode not in 'ucerf_hazard ucerf_risk'):
csm = parallel_split_filter(csm, srcfilter, not oqparam.fast_sampling,
monitor('split_filter'))
return csm
[docs]def split_filter(srcs, srcfilter, seed, monitor):
"""
Split the given source and filter the subsources by distance and by
magnitude. Perform sampling if a nontrivial sample_factor is passed.
Yields a pair (split_sources, split_time) if split_sources is non-empty.
"""
splits, stime = split_sources(srcs)
if splits and seed:
# debugging tip to reduce the size of a calculation
splits = random_filtered_sources(splits, srcfilter, seed)
# NB: for performance, sample before splitting
if splits and srcfilter:
splits = list(srcfilter.filter(splits))
if splits:
yield splits, stime
[docs]def only_filter(srcs, srcfilter, dummy, monitor):
"""
Filter the given sources. Yield a pair (filtered_sources, {src.id: 0})
if there are filtered sources.
"""
srcs = list(srcfilter.filter(srcs))
if srcs:
yield srcs, {src.id: 0 for src in srcs}
[docs]def parallel_split_filter(csm, srcfilter, split, monitor):
"""
Apply :func:`split_filter` in parallel to the composite source model.
:returns: a new :class:`openquake.commonlib.source.CompositeSourceModel`
"""
mon = monitor('split_filter')
seed = int(os.environ.get('OQ_SAMPLE_SOURCES', 0))
msg = 'Splitting/filtering' if split else 'Filtering'
logging.info('%s sources with %s', msg, srcfilter.__class__.__name__)
sources = csm.get_sources()
dist = 'no' if os.environ.get('OQ_DISTRIBUTE') == 'no' else 'processpool'
smap = parallel.Starmap.apply(
split_filter if split else only_filter,
(sources, srcfilter, seed, mon),
maxweight=RUPTURES_PER_BLOCK, distribute=dist,
progress=logging.debug, weight=operator.attrgetter('num_ruptures'))
if monitor.hdf5:
source_info = monitor.hdf5['source_info']
source_info.attrs['has_dupl_sources'] = csm.has_dupl_sources
srcs_by_grp = collections.defaultdict(list)
arr = numpy.zeros((len(sources), 2), F32)
for splits, stime in smap:
for split in splits:
i = split.id
arr[i, 0] += stime[i] # split_time
arr[i, 1] += 1 # num_split
srcs_by_grp[split.src_group_id].append(split)
if not srcs_by_grp:
raise RuntimeError('All sources were filtered away!')
elif monitor.hdf5:
source_info[:, 'split_time'] = arr[:, 0]
source_info[:, 'num_split'] = arr[:, 1]
return csm.new(srcs_by_grp)
[docs]def get_imts(oqparam):
"""
Return a sorted list of IMTs as hazardlib objects
"""
return list(map(imt.from_string, sorted(oqparam.imtls)))
[docs]def get_risk_model(oqparam):
"""
Return a :class:`openquake.risklib.riskinput.CompositeRiskModel` instance
:param oqparam:
an :class:`openquake.commonlib.oqvalidation.OqParam` instance
"""
rmdict = get_risk_models(oqparam)
oqparam.set_risk_imtls(rmdict)
if oqparam.calculation_mode.endswith('_bcr'):
retro = get_risk_models(oqparam, 'vulnerability_retrofitted')
else:
retro = {}
return riskinput.CompositeRiskModel(oqparam, rmdict, retro)
[docs]def get_exposure(oqparam):
"""
Read the full exposure in memory and build a list of
:class:`openquake.risklib.asset.Asset` instances.
:param oqparam:
an :class:`openquake.commonlib.oqvalidation.OqParam` instance
:returns:
an :class:`Exposure` instance or a compatible AssetCollection
"""
exposure = asset.Exposure.read(
oqparam.inputs['exposure'], oqparam.calculation_mode,
oqparam.region, oqparam.ignore_missing_costs,
by_country='country' in oqparam.aggregate_by)
exposure.mesh, exposure.assets_by_site = exposure.get_mesh_assets_by_site()
return exposure
[docs]def get_sitecol_assetcol(oqparam, haz_sitecol=None, cost_types=()):
"""
:param oqparam: calculation parameters
:param haz_sitecol: the hazard site collection
:param cost_types: the expected cost types
:returns: (site collection, asset collection, discarded)
"""
global exposure
if exposure is None:
# haz_sitecol not extracted from the exposure
exposure = get_exposure(oqparam)
if haz_sitecol is None:
haz_sitecol = get_site_collection(oqparam)
missing = set(cost_types) - set(exposure.cost_types['name']) - set(
['occupants']) # TODO: remove occupants and fragility special cases
if missing and not oqparam.calculation_mode.endswith('damage'):
expo = oqparam.inputs.get('exposure', '')
raise InvalidFile(
'Expected cost types %s but the exposure %r contains %s' % (
cost_types, expo, exposure.cost_types['name']))
if oqparam.region_grid_spacing:
haz_distance = oqparam.region_grid_spacing
if haz_distance != oqparam.asset_hazard_distance:
logging.info('Using asset_hazard_distance=%d km instead of %d km',
haz_distance, oqparam.asset_hazard_distance)
else:
haz_distance = oqparam.asset_hazard_distance
if haz_sitecol.mesh != exposure.mesh:
# associate the assets to the hazard sites
tot_assets = sum(len(assets) for assets in exposure.assets_by_site)
sitecol, assets_by, discarded = geo.utils.assoc(
exposure.assets_by_site, haz_sitecol,
oqparam.asset_hazard_distance, 'filter', exposure.asset_refs)
assets_by_site = [[] for _ in sitecol.complete.sids]
num_assets = 0
for sid, assets in zip(sitecol.sids, assets_by):
assets_by_site[sid] = assets
num_assets += len(assets)
logging.info(
'Associated %d assets to %d sites', num_assets, len(sitecol))
if num_assets < tot_assets:
logging.warn('Discarded %d assets outside the '
'asset_hazard_distance of %d km',
tot_assets - num_assets, haz_distance)
else:
# asset sites and hazard sites are the same
sitecol = haz_sitecol
assets_by_site = exposure.assets_by_site
discarded = []
logging.info('Read %d sites and %d assets from the exposure',
len(sitecol), sum(len(a) for a in assets_by_site))
asset_refs = numpy.array(
[exposure.asset_refs[asset.ordinal]
for assets in assets_by_site for asset in assets])
assetcol = asset.AssetCollection(
asset_refs, assets_by_site, exposure.tagcol, exposure.cost_calculator,
oqparam.time_event, exposure.occupancy_periods)
if (not oqparam.hazard_calculation_id and 'gmfs' not in oqparam.inputs
and 'hazard_curves' not in oqparam.inputs
and sitecol is not sitecol.complete):
assetcol = assetcol.reduce_also(sitecol)
return sitecol, assetcol, discarded
def _get_gmfs(oqparam):
M = len(oqparam.imtls)
assert M, ('oqparam.imtls is empty, did you call '
'oqparam.set_risk_imtls(get_risk_models(oqparam))?')
fname = oqparam.inputs['gmfs']
if fname.endswith('.csv'):
array = writers.read_composite_array(fname).array
R = len(numpy.unique(array['rlzi']))
if R > 1:
raise InvalidFile('%s: found %d realizations, currently only one '
'realization is supported' % (fname, R))
# the array has the structure rlzi, sid, eid, gmv_PGA, gmv_...
dtlist = [(name, array.dtype[name]) for name in array.dtype.names[:3]]
required_imts = list(oqparam.imtls)
imts = [name[4:] for name in array.dtype.names[3:]]
if imts != required_imts:
raise ValueError('Required %s, but %s contains %s' % (
required_imts, fname, imts))
dtlist.append(('gmv', (F32, M)))
eids = numpy.unique(array['eid'])
E = len(eids)
found_eids = set(eids)
expected_eids = set(range(E)) # expected incremental eids
missing_eids = expected_eids - found_eids
if missing_eids:
raise InvalidFile('Missing eids in the gmfs.csv file: %s'
% missing_eids)
assert expected_eids == found_eids, (expected_eids, found_eids)
eidx = {eid: e for e, eid in enumerate(eids)}
sitecol = get_site_collection(oqparam)
expected_sids = set(sitecol.sids)
found_sids = set(numpy.unique(array['sid']))
missing_sids = found_sids - expected_sids
if missing_sids:
raise InvalidFile(
'Found site IDs missing in the sites.csv file: %s' %
missing_sids)
N = len(sitecol)
gmfs = numpy.zeros((R, N, E, M), F32)
counter = collections.Counter()
for row in array.view(dtlist):
key = row['rlzi'], row['sid'], eidx[row['eid']]
gmfs[key] = row['gmv']
counter[key] += 1
dupl = [key for key in counter if counter[key] > 1]
if dupl:
raise InvalidFile('Duplicated (rlzi, sid, eid) in the GMFs file: '
'%s' % dupl)
elif fname.endswith('.xml'):
eids, gmfs_by_imt = get_scenario_from_nrml(oqparam, fname)
N, E = gmfs_by_imt.shape
gmfs = numpy.zeros((1, N, E, M), F32)
for imti, imtstr in enumerate(oqparam.imtls):
gmfs[0, :, :, imti] = gmfs_by_imt[imtstr]
else:
raise NotImplemented('Reading from %s' % fname)
return eids, gmfs
[docs]def get_pmap_from_csv(oqparam, fnames):
"""
:param oqparam:
an :class:`openquake.commonlib.oqvalidation.OqParam` instance
:param fnames:
a space-separated list of .csv relative filenames
:returns:
the site mesh and the hazard curves read by the .csv files
"""
if not oqparam.imtls:
oqparam.set_risk_imtls(get_risk_models(oqparam))
if not oqparam.imtls:
raise ValueError('Missing intensity_measure_types_and_levels in %s'
% oqparam.inputs['job_ini'])
dic = {wrapper.imt: wrapper.array
for wrapper in map(writers.read_composite_array, fnames)}
array = dic[next(iter(dic))]
mesh = geo.Mesh(array['lon'], array['lat'])
num_levels = sum(len(imls) for imls in oqparam.imtls.values())
data = numpy.zeros((len(mesh), num_levels))
level = 0
for im in oqparam.imtls:
arr = dic[im]
for poe in arr.dtype.names[3:]:
data[:, level] = arr[poe]
level += 1
for field in ('lon', 'lat', 'depth'): # sanity check
numpy.testing.assert_equal(arr[field], array[field])
return mesh, ProbabilityMap.from_array(data, range(len(mesh)))
[docs]@deprecated('Use the .csv format for the hazard curves instead')
def get_pmap_from_nrml(oqparam, fname):
"""
:param oqparam:
an :class:`openquake.commonlib.oqvalidation.OqParam` instance
:param fname:
an XML file containing hazard curves
:returns:
site mesh, curve array
"""
hcurves_by_imt = {}
oqparam.hazard_imtls = imtls = {}
for hcurves in nrml.read(fname):
imt = hcurves['IMT']
oqparam.investigation_time = hcurves['investigationTime']
if imt == 'SA':
imt += '(%s)' % hcurves['saPeriod']
imtls[imt] = ~hcurves.IMLs
data = sorted((~node.Point.pos, ~node.poEs) for node in hcurves[1:])
hcurves_by_imt[imt] = numpy.array([d[1] for d in data])
lons, lats = [], []
for xy, poes in data:
lons.append(xy[0])
lats.append(xy[1])
mesh = geo.Mesh(numpy.array(lons), numpy.array(lats))
num_levels = sum(len(v) for v in imtls.values())
array = numpy.zeros((len(mesh), num_levels))
imtls = DictArray(imtls)
for imt_ in hcurves_by_imt:
array[:, imtls(imt_)] = hcurves_by_imt[imt_]
return mesh, ProbabilityMap.from_array(array, range(len(mesh)))
# used in get_scenario_from_nrml
def _extract_eids_sitecounts(gmfset):
eids = set()
counter = collections.Counter()
for gmf in gmfset:
eids.add(gmf['ruptureId'])
for node in gmf:
counter[node['lon'], node['lat']] += 1
eids = numpy.array(sorted(eids), numpy.uint64)
if (eids != numpy.arange(len(eids), dtype=numpy.uint64)).any():
raise ValueError('There are ruptureIds in the gmfs_file not in the '
'range [0, %d)' % len(eids))
return eids, counter
[docs]@deprecated('Use the .csv format for the GMFs instead')
def get_scenario_from_nrml(oqparam, fname):
"""
:param oqparam:
an :class:`openquake.commonlib.oqvalidation.OqParam` instance
:param fname:
the NRML files containing the GMFs
:returns:
a pair (eids, gmf array)
"""
if not oqparam.imtls:
oqparam.set_risk_imtls(get_risk_models(oqparam))
imts = sorted(oqparam.imtls)
num_imts = len(imts)
imt_dt = numpy.dtype([(imt, F32) for imt in imts])
gmfset = nrml.read(fname).gmfCollection.gmfSet
eids, sitecounts = _extract_eids_sitecounts(gmfset)
coords = sorted(sitecounts)
oqparam.sites = [(lon, lat, 0) for lon, lat in coords]
site_idx = {lonlat: i for i, lonlat in enumerate(coords)}
oqparam.number_of_ground_motion_fields = num_events = len(eids)
num_sites = len(oqparam.sites)
gmf_by_imt = numpy.zeros((num_events, num_sites), imt_dt)
counts = collections.Counter()
for i, gmf in enumerate(gmfset):
if len(gmf) != num_sites: # there must be one node per site
raise InvalidFile('Expected %d sites, got %d nodes in %s, line %d'
% (num_sites, len(gmf), fname, gmf.lineno))
counts[gmf['ruptureId']] += 1
imt = gmf['IMT']
if imt == 'SA':
imt = 'SA(%s)' % gmf['saPeriod']
for node in gmf:
sid = site_idx[node['lon'], node['lat']]
gmf_by_imt[imt][i % num_events, sid] = node['gmv']
for rupid, count in sorted(counts.items()):
if count < num_imts:
raise InvalidFile("Found a missing ruptureId %d in %s" %
(rupid, fname))
elif count > num_imts:
raise InvalidFile("Found a duplicated ruptureId '%s' in %s" %
(rupid, fname))
expected_gmvs_per_site = num_imts * len(eids)
for lonlat, counts in sitecounts.items():
if counts != expected_gmvs_per_site:
raise InvalidFile(
'%s: expected %d gmvs at location %s, found %d' %
(fname, expected_gmvs_per_site, lonlat, counts))
return eids, gmf_by_imt.T
[docs]def get_mesh_hcurves(oqparam):
"""
Read CSV data in the format `lon lat, v1-vN, w1-wN, ...`.
:param oqparam:
an :class:`openquake.commonlib.oqvalidation.OqParam` instance
:returns:
the mesh of points and the data as a dictionary
imt -> array of curves for each site
"""
imtls = oqparam.imtls
lon_lats = set()
data = AccumDict() # imt -> list of arrays
ncols = len(imtls) + 1 # lon_lat + curve_per_imt ...
csvfile = oqparam.inputs['hazard_curves']
for line, row in enumerate(csv.reader(csvfile), 1):
try:
if len(row) != ncols:
raise ValueError('Expected %d columns, found %d' %
ncols, len(row))
x, y = row[0].split()
lon_lat = valid.longitude(x), valid.latitude(y)
if lon_lat in lon_lats:
raise DuplicatedPoint(lon_lat)
lon_lats.add(lon_lat)
for i, imt_ in enumerate(imtls, 1):
values = valid.decreasing_probabilities(row[i])
if len(values) != len(imtls[imt_]):
raise ValueError('Found %d values, expected %d' %
(len(values), len(imtls([imt_]))))
data += {imt_: [numpy.array(values)]}
except (ValueError, DuplicatedPoint) as err:
raise err.__class__('%s: file %s, line %d' % (err, csvfile, line))
lons, lats = zip(*sorted(lon_lats))
mesh = geo.Mesh(numpy.array(lons), numpy.array(lats))
return mesh, {imt: numpy.array(lst) for imt, lst in data.items()}
# used in utils/reduce_sm and utils/extract_source
[docs]def reduce_source_model(smlt_file, source_ids, remove=True):
"""
Extract sources from the composite source model
"""
found = 0
to_remove = []
for path in logictree.collect_info(smlt_file).smpaths:
root = nrml.read(path)
model = Node('sourceModel', root[0].attrib)
origmodel = root[0]
if root['xmlns'] == 'http://openquake.org/xmlns/nrml/0.4':
for src_node in origmodel:
if src_node['id'] in source_ids:
model.nodes.append(src_node)
else: # nrml/0.5
for src_group in origmodel:
sg = copy.copy(src_group)
sg.nodes = []
weights = src_group.get('srcs_weights')
if weights:
assert len(weights) == len(src_group.nodes)
else:
weights = [1] * len(src_group.nodes)
src_group['srcs_weights'] = reduced_weigths = []
for src_node, weight in zip(src_group, weights):
if src_node['id'] in source_ids:
found += 1
sg.nodes.append(src_node)
reduced_weigths.append(weight)
if sg.nodes:
model.nodes.append(sg)
shutil.copy(path, path + '.bak')
if model:
with open(path, 'wb') as f:
nrml.write([model], f, xmlns=root['xmlns'])
logging.warn('Reduced %s' % path)
elif remove: # remove the files completely reduced
to_remove.append(path)
if found:
for path in to_remove:
os.remove(path)
else:
logging.warn('Sources %s not found', source_ids)
# used in oq zip and oq checksum
def _checksum(fname, checksum):
if not os.path.exists(fname):
zpath = os.path.splitext(fname)[0] + '.zip'
if not os.path.exists(zpath):
raise OSError('No such file: %s or %s' % (fname, zpath))
data = open(zpath, 'rb').read()
else:
data = open(fname, 'rb').read()
return zlib.adler32(data, checksum) & 0xffffffff
[docs]def get_checksum32(oqparam, hazard=False):
"""
Build an unsigned 32 bit integer from the input files of a calculation.
:param oqparam: an OqParam instance
:param hazard: if True, consider only the hazard files
:returns: the checkume
"""
# NB: using adler32 & 0xffffffff is the documented way to get a checksum
# which is the same between Python 2 and Python 3
checksum = 0
for fname in get_input_files(oqparam, hazard):
checksum = _checksum(fname, checksum)
if hazard:
hazard_params = []
for key, val in vars(oqparam).items():
if key in ('rupture_mesh_spacing', 'complex_fault_mesh_spacing',
'width_of_mfd_bin', 'area_source_discretization',
'random_seed', 'ses_seed', 'truncation_level',
'maximum_distance', 'investigation_time',
'number_of_logic_tree_samples', 'imtls',
'ses_per_logic_tree_path', 'minimum_magnitude',
'prefilter_sources', 'sites', 'fast_sampling',
'pointsource_distance', 'filter_distance'):
hazard_params.append('%s = %s' % (key, val))
data = '\n'.join(hazard_params).encode('utf8')
checksum = zlib.adler32(data, checksum) & 0xffffffff
return checksum