# -*- coding: utf-8 -*-
# vim: tabstop=4 shiftwidth=4 softtabstop=4
#
# Copyright (c) 2016-2020 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 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/>.
from openquake.baselib.python3compat import zip
import numpy
F32 = numpy.float32
F64 = numpy.float64
BYTES_PER_FLOAT = 8
[docs]class AllEmptyProbabilityMaps(ValueError):
"""
Raised by get_shape(pmaps) if all passed probability maps are empty
"""
[docs]class ProbabilityCurve(object):
"""
This class is a small wrapper over an array of PoEs associated to
a set of intensity measure types and levels. It provides a few operators,
including the complement operator `~`
~p = 1 - p
and the inclusive or operator `|`
p = p1 | p2 = ~(~p1 * ~p2)
Such operators are implemented efficiently at the numpy level, by
dispatching on the underlying array.
Here is an example of use:
>>> poe = ProbabilityCurve(numpy.array([0.1, 0.2, 0.3, 0, 0]))
>>> ~(poe | poe) * .5
<ProbabilityCurve
[0.405 0.32 0.245 0.5 0.5 ]>
"""
def __init__(self, array):
self.array = array
def __or__(self, other):
if other == 0:
return self
else:
return self.__class__(1. - (1. - self.array) * (1. - other.array))
__ror__ = __or__
def __iadd__(self, other):
# this is used when composing mutually exclusive probabilities
self.array += other.array
return self
def __add__(self, other):
# this is used when composing mutually exclusive probabilities
self.array += other.array
return self.__class__(self.array)
def __mul__(self, other):
if isinstance(other, self.__class__):
return self.__class__(self.array * other.array)
elif other == 1:
return self
else:
return self.__class__(self.array * other)
__rmul__ = __mul__
def __pow__(self, n):
return self.__class__(self.array ** n)
def __invert__(self):
return self.__class__(1. - self.array)
def __bool__(self):
return bool(self.array.any())
def __repr__(self):
return '<ProbabilityCurve\n%s>' % self.array
# used when exporting to HDF5
[docs] def convert(self, imtls, idx=0):
"""
Convert a probability curve into a record of dtype `imtls.dt`.
:param imtls: DictArray instance
:param idx: extract the data corresponding to the given inner index
"""
curve = numpy.zeros(1, imtls.dt)
for imt in imtls:
curve[imt] = self.array[imtls(imt), idx]
return curve[0]
[docs]class ProbabilityMap(dict):
"""
A dictionary site_id -> ProbabilityCurve. It defines the complement
operator `~`, performing the complement on each curve
~p = 1 - p
and the "inclusive or" operator `|`:
m = m1 | m2 = {sid: m1[sid] | m2[sid] for sid in all_sids}
Such operators are implemented efficiently at the numpy level, by
dispatching on the underlying array. Moreover there is a classmethod
.build(L, I, sids, initvalue) to build initialized instances of
:class:`ProbabilityMap`. The map can be represented as 3D array of shape
(shape_x, shape_y, shape_z) = (N, L, I), where N is the number of site IDs,
L the total number of hazard levels and I the number of GSIMs.
"""
[docs] @classmethod
def build(cls, shape_y, shape_z, sids, initvalue=0., dtype=F64):
"""
:param shape_y: the total number of intensity measure levels
:param shape_z: the number of inner levels
:param sids: a set of site indices
:param initvalue: the initial value of the probability (default 0)
:returns: a ProbabilityMap dictionary
"""
dic = cls(shape_y, shape_z)
for sid in sids:
dic.setdefault(sid, initvalue, dtype)
return dic
[docs] @classmethod
def from_array(cls, array, sids):
"""
:param array: array of shape (N, L) or (N, L, I)
:param sids: array of N site IDs
"""
n_sites = len(sids)
n = len(array)
if n_sites != n:
raise ValueError('Passed %d site IDs, but the array has length %d'
% (n_sites, n))
if len(array.shape) == 2: # shape (N, L) -> (N, L, 1)
array = array.reshape(array.shape + (1,))
self = cls(*array.shape[1:])
for sid, poes in zip(sids, array):
self[sid] = ProbabilityCurve(poes)
return self
def __init__(self, shape_y, shape_z=1):
self.shape_y = shape_y
self.shape_z = shape_z
[docs] def setdefault(self, sid, value, dtype=F64):
"""
Works like `dict.setdefault`: if the `sid` key is missing, it fills
it with an array and returns the associate ProbabilityCurve
:param sid: site ID
:param value: value used to fill the returned ProbabilityCurve
:param dtype: dtype used internally (F32 or F64)
"""
try:
return self[sid]
except KeyError:
array = numpy.empty((self.shape_y, self.shape_z), dtype)
array.fill(value)
pc = ProbabilityCurve(array)
self[sid] = pc
return pc
@property
def sids(self):
"""The ordered keys of the map as a numpy.uint32 array"""
return numpy.array(sorted(self), numpy.uint32)
@property
def array(self):
"""
The underlying array of shape (N, L, I)
"""
return numpy.array([self[sid].array for sid in sorted(self)])
@property
def nbytes(self):
"""The size of the underlying array"""
try:
N, L, I = get_shape([self])
except AllEmptyProbabilityMaps:
return 0
return BYTES_PER_FLOAT * N * L * I
# used when exporting to HDF5
[docs] def convert(self, imtls, nsites, idx=0):
"""
Convert a probability map into a composite array of length `nsites`
and dtype `imtls.dt`.
:param imtls:
DictArray instance
:param nsites:
the total number of sites
:param idx:
index on the z-axis (default 0)
"""
curves = numpy.zeros(nsites, imtls.dt)
for imt in curves.dtype.names:
curves_by_imt = curves[imt]
for sid in self:
curves_by_imt[sid] = self[sid].array[imtls(imt), idx]
return curves
[docs] def filter(self, sids):
"""
Extracs a submap of self for the given sids.
"""
dic = self.__class__(self.shape_y, self.shape_z)
for sid in sids:
try:
dic[sid] = self[sid]
except KeyError:
pass
return dic
def __ior__(self, other):
if not other:
return self
if (other.shape_y, other.shape_z) != (self.shape_y, self.shape_z):
raise ValueError('%s has inconsistent shape with %s' %
(other, self))
self_sids = set(self)
other_sids = set(other)
for sid in self_sids & other_sids:
self[sid] = self[sid] | other[sid]
for sid in other_sids - self_sids:
self[sid] = other[sid]
return self
def __or__(self, other):
new = self.__class__(self.shape_y, self.shape_z)
new.update(self)
new |= other
return new
__ror__ = __or__
def __add__(self, other):
try:
other.get
is_pmap = True
sids = set(self) | set(other)
except AttributeError: # no .get method, assume a float
is_pmap = False
assert 0. <= other <= 1., other # must be a probability
sids = set(self)
new = self.__class__(self.shape_y, self.shape_z)
for sid in sids:
prob = other.get(sid, 1) if is_pmap else other
new[sid] = self.get(sid, 1) + prob
return new
def __iadd__(self, other):
# this is used when composing mutually exclusive probabilities
for sid in other:
pcurve = self.setdefault(sid, 0)
pcurve += other[sid]
return self
def __mul__(self, other):
try:
other.get
is_pmap = True
sids = set(self) | set(other)
except AttributeError: # no .get method, assume a float
is_pmap = False
assert 0. <= other <= 1., other # must be a probability
sids = set(self)
new = self.__class__(self.shape_y, self.shape_z)
for sid in sids:
prob = other.get(sid, 1) if is_pmap else other
new[sid] = self.get(sid, 1) * prob
return new
def __ipow__(self, n):
for sid, pcurve in self.items():
self[sid] = pcurve ** n
return self
def __pow__(self, n):
new = self.__class__(self.shape_y, self.shape_z)
for sid, pcurve in self.items():
new[sid] = pcurve ** n
return new
def __invert__(self):
new = self.__class__(self.shape_y, self.shape_z)
for sid in self:
if (self[sid].array != 1.).any():
new[sid] = ~self[sid] # store only nonzero probabilities
return new
def __getstate__(self):
return dict(shape_y=self.shape_y, shape_z=self.shape_z)
def __toh5__(self):
# converts to an array of shape (num_sids, shape_y, shape_z)
size = len(self)
sids = self.sids
shape = (size, self.shape_y, self.shape_z)
array = numpy.zeros(shape, F64)
for i, sid in numpy.ndenumerate(sids):
try:
array[i] = self[sid].array
except ValueError as exc:
# this should never happen, but I got a could not broadcast
# input array once for Australia
raise ValueError('%s on site_id=%d' % (exc, sid))
return dict(array=array, sids=sids), {}
def __fromh5__(self, dic, attrs):
# rebuild the map from sids and probs arrays
array = dic['array']
sids = dic['sids']
self.shape_y = array.shape[1]
self.shape_z = array.shape[2]
for sid, prob in zip(sids, array):
self[sid] = ProbabilityCurve(prob)
def __repr__(self):
return '<%s %d, %d, %d>' % (self.__class__.__name__, len(self),
self.shape_y, self.shape_z)
[docs]def get_shape(pmaps):
"""
:param pmaps: a set of homogenous ProbabilityMaps
:returns: the common shape (N, L, I)
"""
for pmap in pmaps:
if pmap:
sid = next(iter(pmap))
break
else:
raise AllEmptyProbabilityMaps(pmaps)
return (len(pmap),) + pmap[sid].array.shape
[docs]def combine(pmaps):
"""
:param pmaps: a set of homogenous ProbabilityMaps
:returns: the combined map
"""
shape = get_shape(pmaps)
res = ProbabilityMap(shape[1], shape[2])
for pmap in pmaps:
res |= pmap
return res