import logging
from . import modifier
from ..paramsets import constrained_by_poisson
from .. import get_backend, default_backend, events
log = logging.getLogger(__name__)
[docs]@modifier(
name='shapesys', constrained=True, pdf_type='poisson', op_code='multiplication'
)
class shapesys(object):
[docs] @classmethod
def required_parset(cls, n_parameters):
return {
'paramset_type': constrained_by_poisson,
'n_parameters': n_parameters,
'modifier': cls.__name__,
'is_constrained': cls.is_constrained,
'is_shared': False,
'inits': (1.0,) * n_parameters,
'bounds': ((1e-10, 10.0),) * n_parameters,
# nb: auxdata/factors set by finalize. Set to non-numeric to crash
# if we fail to set auxdata/factors correctly
'auxdata': (None,) * n_parameters,
'factors': (None,) * n_parameters,
}
class shapesys_combined(object):
def __init__(self, shapesys_mods, pdfconfig, mega_mods):
pnames = [pname for pname, _ in shapesys_mods]
keys = ['{}/{}'.format(mtype, m) for m, mtype in shapesys_mods]
shapesys_mods = [m for m, _ in shapesys_mods]
self._shapesys_mods = shapesys_mods
self._pnames = pnames
self._parindices = list(range(len(pdfconfig.suggested_init())))
self._shapesys_indices = [
self._parindices[pdfconfig.par_slice(p)] for p in pnames
]
self._shapesys_mask = [
[[mega_mods[s][m]['data']['mask']] for s in pdfconfig.samples] for m in keys
]
self.__shapesys_uncrt = default_backend.astensor(
[
[
[
mega_mods[s][m]['data']['uncrt'],
mega_mods[s][m]['data']['nom_data'],
]
for s in pdfconfig.samples
]
for m in keys
]
)
if self._shapesys_indices:
access_rows = []
shapesys_mask = default_backend.astensor(self._shapesys_mask)
for mask, inds in zip(shapesys_mask, self._shapesys_indices):
summed_mask = default_backend.sum(mask[:, 0, :], axis=0)
assert default_backend.shape(
summed_mask[summed_mask > 0]
) == default_backend.shape(default_backend.astensor(inds))
# make masks of > 0 and == 0
positive_mask = summed_mask > 0
zero_mask = summed_mask == 0
# then apply the mask
summed_mask[positive_mask] = inds
summed_mask[zero_mask] = len(self._parindices) - 1
access_rows.append(summed_mask.tolist())
self._factor_access_indices = default_backend.tolist(
default_backend.stack(access_rows)
)
self.finalize(pdfconfig)
else:
self._factor_access_indices = None
self._precompute()
events.subscribe('tensorlib_changed')(self._precompute)
def _precompute(self):
tensorlib, _ = get_backend()
self.shapesys_mask = tensorlib.astensor(self._shapesys_mask)
self.shapesys_default = tensorlib.ones(tensorlib.shape(self.shapesys_mask))
if self._shapesys_indices:
self.factor_access_indices = tensorlib.astensor(
self._factor_access_indices, dtype='int'
)
self.default_value = tensorlib.astensor([1.0])
self.sample_ones = tensorlib.ones(tensorlib.shape(self.shapesys_mask)[1])
self.alpha_ones = tensorlib.astensor([1])
else:
self.factor_access_indices = None
def finalize(self, pdfconfig):
for uncert_this_mod, pname in zip(self.__shapesys_uncrt, self._pnames):
unc_nom = default_backend.astensor(
[x for x in uncert_this_mod[:, :, :] if any(x[0][x[0] > 0])]
)
unc = unc_nom[0, 0]
nom = unc_nom[0, 1]
unc_sq = default_backend.power(unc, 2)
nom_sq = default_backend.power(nom, 2)
# the below tries to filter cases in which
# this modifier is not used by checking non
# zeroness.. shoudl probably use mask
numerator = default_backend.where(
unc_sq > 0, nom_sq, default_backend.zeros(unc_sq.shape)
)
denominator = default_backend.where(
unc_sq > 0, unc_sq, default_backend.ones(unc_sq.shape)
)
factors = numerator / denominator
factors = factors[factors > 0]
assert len(factors) == pdfconfig.param_set(pname).n_parameters
pdfconfig.param_set(pname).factors = default_backend.tolist(factors)
pdfconfig.param_set(pname).auxdata = default_backend.tolist(factors)
def apply(self, pars):
tensorlib, _ = get_backend()
if self.factor_access_indices is None:
return
tensorlib, _ = get_backend()
factor_row = tensorlib.gather(
tensorlib.concatenate([tensorlib.astensor(pars), self.default_value]),
self.factor_access_indices,
)
results_shapesys = tensorlib.einsum(
's,a,mb->msab',
tensorlib.astensor(self.sample_ones),
tensorlib.astensor(self.alpha_ones),
factor_row,
)
results_shapesys = tensorlib.where(
self.shapesys_mask, results_shapesys, self.shapesys_default
)
return results_shapesys