Source code for openquake.calculators.classical_risk

# -*- coding: utf-8 -*-
# vim: tabstop=4 shiftwidth=4 softtabstop=4
#
# Copyright (C) 2014-2020 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 numpy
from openquake.baselib.python3compat import encode
from openquake.hazardlib.stats import compute_stats
from openquake.risklib import scientific
from openquake.calculators import base


F32 = numpy.float32


[docs]def classical_risk(riskinputs, crmodel, param, monitor): """ Compute and return the average losses for each asset. :param riskinputs: :class:`openquake.risklib.riskinput.RiskInput` objects :param crmodel: a :class:`openquake.risklib.riskinput.CompositeRiskModel` instance :param param: dictionary of extra parameters :param monitor: :class:`openquake.baselib.performance.Monitor` instance """ result = dict(loss_curves=[], stat_curves=[]) weights = [w['default'] for w in param['weights']] statnames, stats = zip(*param['stats']) for ri in riskinputs: A = len(ri.assets) L = len(crmodel.lti) R = ri.hazard_getter.num_rlzs loss_curves = numpy.zeros((R, L, A), object) avg_losses = numpy.zeros((R, L, A)) for out in ri.gen_outputs(crmodel, monitor): r = out.rlzi for l, loss_type in enumerate(crmodel.loss_types): # loss_curves has shape (A, C) for i, asset in enumerate(ri.assets): loss_curves[out.rlzi, l, i] = lc = out[loss_type][i] aid = asset['ordinal'] avg = scientific.average_loss(lc) avg_losses[r, l, i] = avg lcurve = (lc['loss'], lc['poe'], avg) result['loss_curves'].append((l, r, aid, lcurve)) # compute statistics for l, loss_type in enumerate(crmodel.loss_types): for i, asset in enumerate(ri.assets): avg_stats = compute_stats(avg_losses[:, l, i], stats, weights) losses = loss_curves[0, l, i]['loss'] all_poes = numpy.array( [loss_curves[r, l, i]['poe'] for r in range(R)]) poes_stats = compute_stats(all_poes, stats, weights) result['stat_curves'].append( (l, asset['ordinal'], losses, poes_stats, avg_stats)) if R == 1: # the realization is the same as the mean del result['loss_curves'] return result
[docs]@base.calculators.add('classical_risk') class ClassicalRiskCalculator(base.RiskCalculator): """ Classical Risk calculator """ core_task = classical_risk precalc = 'classical' accept_precalc = ['classical']
[docs] def pre_execute(self): """ Associate the assets to the sites and build the riskinputs. """ oq = self.oqparam super().pre_execute() if 'poes' not in self.datastore: # when building short report return full_lt = self.datastore['full_lt'] self.realizations = full_lt.get_realizations() weights = [rlz.weight for rlz in self.realizations] stats = list(oq.hazard_stats().items()) self.param = dict(stats=stats, weights=weights) self.riskinputs = self.build_riskinputs('poe') self.A = len(self.assetcol) self.L = len(self.crmodel.loss_types) self.S = len(oq.hazard_stats())
[docs] def post_execute(self, result): """ Saving loss curves in the datastore. :param result: aggregated result of the task classical_risk """ curve_res = {cp.loss_type: cp.curve_resolution for cp in self.crmodel.curve_params if cp.user_provided} self.loss_curve_dt = scientific.build_loss_curve_dt( curve_res, insured_losses=False) ltypes = self.crmodel.loss_types # loss curves stats are generated always stats = encode(list(self.oqparam.hazard_stats())) stat_curves = numpy.zeros((self.A, self.S), self.loss_curve_dt) avg_losses = numpy.zeros((self.A, self.S, self.L), F32) for l, a, losses, statpoes, statloss in result['stat_curves']: stat_curves_lt = stat_curves[ltypes[l]] for s in range(self.S): avg_losses[a, s, l] = statloss[s] base.set_array(stat_curves_lt['poes'][a, s], statpoes[s]) base.set_array(stat_curves_lt['losses'][a, s], losses) self.datastore['avg_losses-stats'] = avg_losses self.datastore.set_attrs('avg_losses-stats', stats=stats) self.datastore['loss_curves-stats'] = stat_curves self.datastore.set_attrs('loss_curves-stats', stats=stats) if self.R > 1: # individual realizations saved only if many loss_curves = numpy.zeros((self.A, self.R), self.loss_curve_dt) avg_losses = numpy.zeros((self.A, self.R, self.L), F32) for l, r, a, (losses, poes, avg) in result['loss_curves']: lc = loss_curves[a, r][ltypes[l]] avg_losses[a, r, l] = avg base.set_array(lc['losses'], losses) base.set_array(lc['poes'], poes) self.datastore['avg_losses-rlzs'] = avg_losses self.datastore['loss_curves-rlzs'] = loss_curves