def view_delta_loss(token, dstore):
"""
Estimate the stocastic error on the loss curve by splitting the events
in odd and even. Example:
$ oq show delta_loss # consider the first loss type
"""
if ':' in token:
_, li = token.split(':')
li = int(li)
else:
li = 0
oq = dstore['oqparam']
efftime = oq.investigation_time * oq.ses_per_logic_tree_path * len(
dstore['weights'])
num_events = len(dstore['events'])
num_events0 = num_events // 2 + (num_events % 2)
num_events1 = num_events // 2
periods = return_periods(efftime, num_events)[1:-1]
K = dstore['risk_by_event'].attrs.get('K', 0)
df = dstore.read_df('risk_by_event', 'event_id', dict(agg_id=K,
loss_id=li))
if len(df) == 0: # for instance no fatalities
return {'delta': numpy.zeros(1)}
mod2 = df.index % 2
losses0 = df['loss'][mod2 == 0]
losses1 = df['loss'][mod2 == 1]
c0 = losses_by_period(losses0, periods, num_events0, efftime / 2)
c1 = losses_by_period(losses1, periods, num_events1, efftime / 2)
dic = dict(loss=losses_by_period(df['loss'], periods, num_events, efftime),
even=c0,
odd=c1,
delta=numpy.abs(c0 - c1) / (c0 + c1))
return pandas.DataFrame(dic, periods)
def post_execute(self, dummy):
"""
Compute and store average losses from the losses_by_event dataset,
and then loss curves and maps.
"""
oq = self.oqparam
if oq.avg_losses:
self.datastore['avg_losses-stats'].attrs['stats'] = [b'mean']
logging.info('Building loss tables')
build_loss_tables(self.datastore)
self.datastore.flush() # just to be sure
shp = self.get_shape(self.L) # (L, T...)
text = ' x '.join('%d(%s)' % (n, t)
for t, n in zip(oq.aggregate_by, shp[1:]))
logging.info('Producing %d(loss_types) x %s loss curves', self.L, text)
builder = get_loss_builder(self.datastore)
self.build_datasets(builder)
self.datastore.swmr_on()
args = [(self.datastore.filename, builder, oq.ses_ratio, rlzi)
for rlzi in range(self.R)]
acc = list(parallel.Starmap(postprocess, args, h5=self.datastore.hdf5))
for r, (curves, maps), agg_losses in acc:
if len(curves): # some realization can give zero contribution
self.datastore['agg_curves-rlzs'][:, r] = curves
if len(maps): # conditional_loss_poes can be empty
self.datastore['agg_maps-rlzs'][:, r] = maps
self.datastore['agg_losses-rlzs'][:, r] = agg_losses
if self.R > 1:
logging.info('Computing aggregate statistics')
set_rlzs_stats(self.datastore, 'agg_curves')
set_rlzs_stats(self.datastore, 'agg_losses')
if oq.conditional_loss_poes:
set_rlzs_stats(self.datastore, 'agg_maps')
# sanity check with the asset_loss_table
if oq.asset_loss_table and len(oq.aggregate_by) == 1:
alt = self.datastore['asset_loss_table'][()]
if alt.sum() == 0: # nothing was saved
return
logging.info('Checking the loss curves')
tags = getattr(self.assetcol.tagcol, oq.aggregate_by[0])[1:]
T = len(tags)
P = len(builder.return_periods)
# sanity check on the loss curves for simple tag aggregation
arr = self.assetcol.aggregate_by(oq.aggregate_by, alt)
# shape (T, E, L)
rlzs = self.datastore['events']['rlz_id']
curves = numpy.zeros((P, self.R, self.L, T))
for t in range(T):
for r in range(self.R):
for l in range(self.L):
curves[:, r, l,
t] = losses_by_period(arr[t, rlzs == r, l],
builder.return_periods,
builder.num_events[r],
builder.eff_time)
numpy.testing.assert_allclose(
curves, self.datastore['agg_curves-rlzs'][()])
def test(self):
# testing convergency of the mean curve
periods = [10, 20, 50, 100, 150, 200, 250]
eff_time = 500
losses = 10**numpy.random.default_rng(42).random(2000)
losses0 = losses[:1000]
losses1 = losses[1000:]
curve0 = scientific.losses_by_period(losses0,
periods,
eff_time=eff_time)
curve1 = scientific.losses_by_period(losses1,
periods,
eff_time=eff_time)
mean = (curve0 + curve1) / 2
full = scientific.losses_by_period(losses,
periods,
eff_time=2 * eff_time)
aae(mean, full, rtol=1E-2) # converges only at 1%
def post_execute(self, times):
"""
Compute and store average losses from the losses_by_event dataset,
and then loss curves and maps.
"""
if len(times):
self.datastore.set_attrs(
'task_info/start_ebrisk', times=times,
events_per_sid=numpy.mean(self.events_per_sid))
oq = self.oqparam
shp = self.get_shape(self.L) # (L, T...)
text = ' x '.join(
'%d(%s)' % (n, t) for t, n in zip(oq.aggregate_by, shp[1:]))
logging.info('Producing %d(loss_types) x %s loss curves', self.L, text)
builder = get_loss_builder(self.datastore)
self.build_datasets(builder)
self.datastore.close()
if 'losses_by_event' in self.datastore.parent:
dstore = self.datastore.parent
else:
dstore = self.datastore
allargs = [(dstore.filename, builder, rlzi) for rlzi in range(self.R)]
mon = performance.Monitor(hdf5=hdf5.File(self.datastore.hdf5cache()))
acc = list(parallel.Starmap(compute_loss_curves_maps, allargs, mon))
# copy performance information from the cache to the datastore
pd = mon.hdf5['performance_data'][()]
hdf5.extend3(self.datastore.filename, 'performance_data', pd)
self.datastore.open('r+') # reopen
self.datastore['task_info/compute_loss_curves_and_maps'] = (
mon.hdf5['task_info/compute_loss_curves_maps'][()])
self.datastore.open('r+')
with self.monitor('saving loss_curves and maps', autoflush=True):
for r, (curves, maps) in acc:
if len(curves): # some realization can give zero contribution
self.datastore['agg_curves-rlzs'][:, r] = curves
if len(maps): # conditional_loss_poes can be empty
self.datastore['agg_maps-rlzs'][:, r] = maps
if self.R > 1:
logging.info('Computing aggregate loss curves statistics')
set_rlzs_stats(self.datastore, 'agg_curves')
self.datastore.set_attrs(
'agg_curves-stats', return_periods=builder.return_periods,
loss_types=' '.join(self.riskmodel.loss_types))
if oq.conditional_loss_poes:
logging.info('Computing aggregate loss maps statistics')
set_rlzs_stats(self.datastore, 'agg_maps')
# sanity check with the asset_loss_table
if oq.asset_loss_table and len(oq.aggregate_by) == 1:
alt = self.datastore['asset_loss_table'][()]
if alt.sum() == 0: # nothing was saved
return
logging.info('Checking the loss curves')
tags = getattr(self.assetcol.tagcol, oq.aggregate_by[0])[1:]
T = len(tags)
P = len(builder.return_periods)
# sanity check on the loss curves for simple tag aggregation
arr = self.assetcol.aggregate_by(oq.aggregate_by, alt)
# shape (T, E, L)
rlzs = self.datastore['events']['rlz']
curves = numpy.zeros((P, self.R, self.L, T))
for t in range(T):
for r in range(self.R):
for l in range(self.L):
curves[:, r, l, t] = losses_by_period(
arr[t, rlzs == r, l],
builder.return_periods,
builder.num_events[r],
builder.eff_time)
numpy.testing.assert_allclose(
curves, self.datastore['agg_curves-rlzs'][()])
def post_execute(self, times):
"""
Compute and store average losses from the losses_by_event dataset,
and then loss curves and maps.
"""
if len(times):
self.datastore.set_attrs('task_info/start_ebrisk',
times=times,
events_per_sid=numpy.mean(
self.events_per_sid))
oq = self.oqparam
shp = self.get_shape(self.L) # (L, T...)
text = ' x '.join('%d(%s)' % (n, t)
for t, n in zip(oq.aggregate_by, shp[1:]))
logging.info('Producing %d(loss_types) x %s loss curves', self.L, text)
builder = get_loss_builder(self.datastore)
self.build_datasets(builder)
self.datastore.close()
if 'losses_by_event' in self.datastore.parent:
dstore = self.datastore.parent
else:
dstore = self.datastore
allargs = [(dstore.filename, builder, rlzi) for rlzi in range(self.R)]
mon = performance.Monitor(hdf5=hdf5.File(self.datastore.hdf5cache()))
acc = list(parallel.Starmap(compute_loss_curves_maps, allargs, mon))
# copy performance information from the cache to the datastore
pd = mon.hdf5['performance_data'][()]
hdf5.extend3(self.datastore.filename, 'performance_data', pd)
self.datastore.open('r+') # reopen
self.datastore['task_info/compute_loss_curves_and_maps'] = (
mon.hdf5['task_info/compute_loss_curves_maps'][()])
self.datastore.open('r+')
with self.monitor('saving loss_curves and maps', autoflush=True):
for r, (curves, maps) in acc:
if len(curves): # some realization can give zero contribution
self.datastore['agg_curves-rlzs'][:, r] = curves
if len(maps): # conditional_loss_poes can be empty
self.datastore['agg_maps-rlzs'][:, r] = maps
if self.R > 1:
logging.info('Computing aggregate loss curves statistics')
set_rlzs_stats(self.datastore, 'agg_curves')
self.datastore.set_attrs('agg_curves-stats',
return_periods=builder.return_periods,
loss_types=' '.join(
self.riskmodel.loss_types))
if oq.conditional_loss_poes:
logging.info('Computing aggregate loss maps statistics')
set_rlzs_stats(self.datastore, 'agg_maps')
# sanity check with the asset_loss_table
if oq.asset_loss_table and len(oq.aggregate_by) == 1:
alt = self.datastore['asset_loss_table'][()]
if alt.sum() == 0: # nothing was saved
return
logging.info('Checking the loss curves')
tags = getattr(self.assetcol.tagcol, oq.aggregate_by[0])[1:]
T = len(tags)
P = len(builder.return_periods)
# sanity check on the loss curves for simple tag aggregation
arr = self.assetcol.aggregate_by(oq.aggregate_by, alt)
# shape (T, E, L)
rlzs = self.datastore['events']['rlz']
curves = numpy.zeros((P, self.R, self.L, T))
for t in range(T):
for r in range(self.R):
for l in range(self.L):
curves[:, r, l,
t] = losses_by_period(arr[t, rlzs == r, l],
builder.return_periods,
builder.num_events[r],
builder.eff_time)
numpy.testing.assert_allclose(
curves, self.datastore['agg_curves-rlzs'][()])
