def calculate_fit_coeffs(self):
"""
Calculate the fit coefficients for each systematic, flavint, bin
for a polynomial.
"""
this_hash = hash_obj(
[self.fit_binning.hash, self.weight_hash] +
[self.params[name].value for name in self.fit_params],
full_hash=self.full_hash
)
if self.fitcoeffs_hash == this_hash:
return self._fit_coeffs
if self.neutrinos:
nu_params = self.nu_params
else:
nu_params = None
if self.muons:
mu_params = self.mu_params
else:
mu_params = None
if self.params['cache_fit'].value:
this_cache_hash = hash_obj(
[self._data.metadata['name'], self._data.metadata['sample'],
self._data.metadata['cuts'], self.fit_binning.hash] +
[self.params[name].value for name in self.fit_params],
full_hash=self.full_hash
)
if self.fitcoeffs_cache_hash == this_cache_hash:
fit_coeffs = deepcopy(self._cached_fc)
elif this_cache_hash in self.disk_cache:
logging.info('Loading fit coefficients from cache.')
self._cached_fc = self.disk_cache[this_cache_hash]
fit_coeffs = deepcopy(self._cached_fc)
self.fitcoeffs_cache_hash = this_cache_hash
else:
fit_coeffs = self._calculate_fit_coeffs(
self._data, ParamSet(p for p in self.params
if p.name in self.fit_params),
self.fit_binning, nu_params, mu_params
)
else:
fit_coeffs = self._calculate_fit_coeffs(
self._data, ParamSet(p for p in self.params
if p.name in self.fit_params),
self.fit_binning, nu_params, mu_params
)
if self.params['cache_fit'].value:
if this_cache_hash not in self.disk_cache:
logging.info('Caching fit coefficients values to disk.')
self.disk_cache[this_cache_hash] = fit_coeffs
self.fitcoeffs_hash = this_hash
self._fit_coeffs = fit_coeffs
return fit_coeffs
class TEST_CONFIGS(object):
"""Default configurations for stages in a minimal test pipeline."""
pipe_cfg = OrderedDict(
pipeline={
"name": "muons",
"output_binning": TEST_BINNING,
"output_key": ("weights"),
"detector_name": None,
}
)
event_generator_cfg = {
"calc_mode": "events",
"apply_mode": "events",
"output_names": ["muon"],
"params": ParamSet(
[
Param(name="n_events", value=1e3, **PARAM_DEFAULTS),
Param(name="seed", value=0, **PARAM_DEFAULTS),
Param(name="random", value=False, **PARAM_DEFAULTS),
]
),
}
aeff_cfg = {
"calc_mode": "events",
"apply_mode": "events",
"params": ParamSet(
[
Param(name="livetime", value=12345 * ureg["seconds"], **PARAM_DEFAULTS),
Param(name="weight_scale", value=1.0, **PARAM_DEFAULTS),
]
),
}
set_variance_cfg = {
"calc_mode": TEST_BINNING,
"apply_mode": TEST_BINNING,
"divide_total_mc": True,
# expected number of unweighted MC events including events that fall outside of
# the analysis binning
"expected_total_mc": 1000,
"variance_scale": 0.1,
}
fix_error_cfg = {
"calc_mode": TEST_BINNING,
"apply_mode": TEST_BINNING,
}
kde_cfg = {
"calc_mode": "events",
"apply_mode": TEST_BINNING,
"bootstrap": False,
"bootstrap_seed": 0,
"bootstrap_niter": 6,
"linearize_log_dims": True,
"stash_hists": False,
"coszen_name": "true_coszen",
"stack_pid": False,
"oversample": 1,
}
def params(self):
"""Returns a ParamSet including all params of all detectors. First the shared params
(if there are some), then all the "single detector" params. If two detectors use a
parameter with the same name (but not shared), the name of the detector is added to the
parameter name (except for the first detector).
"""
params = ParamSet()
for p_name in self.shared_params:
for distribution_maker in self:
try:
params.extend(distribution_maker.params[p_name])
break # shared param found, can continue with the next shared param
except:
continue # shared param was not in this DistributionMaker, so search in the next one
for distribution_maker in self:
for param in distribution_maker.params:
if param.name in params.names and param.name in self.shared_params:
continue # shared param is already in param set, can continue with the next param
elif param.name in params.names: # two parameters with the same name but not shared
# add detector name to the parameter name
changed_param = deepcopy(param)
changed_param.name = param.name + '_' + distribution_maker._detector_name
params.extend(changed_param)
else:
params.extend(param)
return params
def update_params(self, params):
for distribution_maker in self:
distribution_maker.update_params(params)
#if None in self.det_names: return # No detector names
if isinstance(params,Param): params = ParamSet(params) # just for the following
for p in params.names: # now update params with det_names inside
for i, det_name in enumerate(self.det_names):
if det_name in p:
cp = deepcopy(params[p])
cp.name = cp.name.replace('_'+det_name, "")
self._distribution_makers[i].update_params(cp)
def create_mc_template(toymc_params, config_file=None, seed=None):
'''
Create MC template out of a pisa pipeline
'''
if seed is not None:
np.random.seed(seed)
Config = parse_pipeline_config(config_file)
new_n_events_data = Param(name='n_events_data',
value=toymc_params.n_data,
prior=None,
range=None,
is_fixed=True)
new_sig_frac = Param(name='signal_fraction',
value=toymc_params.signal_fraction,
prior=None,
range=None,
is_fixed=True)
new_stats_factor = Param(name='stats_factor',
value=toymc_params.stats_factor,
prior=None,
range=None,
is_fixed=True)
# These should match the values of the config file, but we override them just in case we need to change these later
new_mu = Param(name='mu',
value=toymc_params.mu,
prior=None,
range=[0, 100],
is_fixed=False)
new_sigma = Param(name='sigma',
value=toymc_params.sigma,
prior=None,
range=None,
is_fixed=True)
Config[('data', 'pi_simple_signal')]['params'].update(p=ParamSet([
new_n_events_data, new_sig_frac, new_stats_factor, new_mu, new_sigma
]))
MCtemplate = DistributionMaker(Config)
return MCtemplate
def pipeline_cfg_from_states(state_dict):
"""Recover a pipeline cfg containing PISA objects from a raw state.
When a pipeline configuration is stored to JSON, the PISA objects turn into
their serialized states. This function looks through the dictionary returned by
`from_json` and recovers the PISA objects such as `ParamSet` and `MultiDimBinning`.
It should really become part of PISA file I/O functionality to read and write
PISA objects inside dictionaries/lists into a JSON and be able to recover
them...
"""
# TODO: Make this a core functionality of PISA
# This is just a mess... some objects have a `from_state` method, some take the
# unpacked state dict as input, some take the state...
pipeline_cfg = collections.OrderedDict()
for stage_key in state_dict.keys():
# need to check all of this manually... no automatic way to do it :(
if stage_key == "pipeline":
pipeline_cfg[stage_key] = copy.deepcopy(state_dict[stage_key])
pipeline_cfg[stage_key]["output_key"] = tuple(
pipeline_cfg[stage_key]["output_key"])
binning_state = pipeline_cfg[stage_key]["output_binning"]
pipeline_cfg[stage_key]["output_binning"] = MultiDimBinning(
**binning_state)
continue
# undo what we did in `serialize_pipeline_cfg` by splitting the keys into tuples
tuple_key = tuple(stage_key.split("__"))
pipeline_cfg[tuple_key] = copy.deepcopy(state_dict[stage_key])
for k in ["calc_mode", "apply_mode", "node_mode"]:
if k in pipeline_cfg[tuple_key]:
if isinstance(pipeline_cfg[tuple_key][k],
collections.abc.Mapping):
pipeline_cfg[tuple_key][k] = MultiDimBinning(
**pipeline_cfg[tuple_key][k])
if "params" in pipeline_cfg[tuple_key].keys():
pipeline_cfg[tuple_key]["params"] = ParamSet(
pipeline_cfg[tuple_key]["params"])
# if any stage takes any other arguments that we didn't think of here, they
# won't work
return pipeline_cfg
def params(self):
params = ParamSet()
for pipeline in self:
params.extend(pipeline.params)
return params
def params(self):
"""pisa.core.param.ParamSet : pipeline's parameters"""
params = ParamSet()
for stage in self:
params.extend(stage.params)
return params
def main():
global SIGMA
args = vars(parse_args())
set_verbosity(args.pop('v'))
center_zero = args.pop('center_zero')
make_pdf = False
if args['pdf']:
make_pdf = True
args['pdf'] = False
outdir = args.pop('outdir')
fileio.mkdir(outdir, mode=0755)
SIGMA *= args.pop('sigma')
cfx_pipe = Pipeline(args.pop('cfx_pipeline'))
signal = args.pop('signal').replace(' ', '').split(',')
output_str = []
for name in signal:
if 'muons' in name or 'noise' in name:
raise AssertionError('Are you trying to unfold muons/noise?')
elif 'all_nu' in name:
output_str = [str(NuFlavIntGroup(f)) for f in ALL_NUFLAVINTS]
else:
output_str.append(NuFlavIntGroup(name))
output_str = [str(f) for f in output_str]
cfx_pipe._output_names = output_str
# Turn off stat fluctuations
stat_param = cfx_pipe.params['stat_fluctuations']
stat_param.value = 0 * ureg.dimensionless
cfx_pipe.update_params(stat_param)
# Get nominal Map
re_param = cfx_pipe.params['regularisation']
re_param.value = 0 * ureg.dimensionless
cfx_pipe.update_params(re_param)
nom_out = cfx_pipe.get_outputs()
re_param.reset()
cfx_pipe.update_params(re_param)
params = ParamSet()
for param in cfx_pipe.params:
if param.name != 'dataset':
params.extend(param)
free = params.free
logging.info('Free params = {0}'.format(free))
contin = True
for f in free:
if 'hole_ice' not in f.name and 'dom_eff' not in f.name:
continue
# if 'atm_muon_scale' in f.name:
# contin = False
# if contin:
# continue
logging.info('Working on parameter {0}'.format(f.name))
if f.prior.kind != 'uniform':
# Use deltaLLH = SIGMA to define +/- sigma for non-uniform
scan_over = np.linspace(*f.range, num=1000) * f.range[0].u
llh = f.prior.llh(scan_over)
dllh = llh - np.min(-llh)
mllh_idx = np.argmin(-llh)
if mllh_idx == 0:
l_sig_idx = 0
else:
l_sig_idx = np.argmin(np.abs(dllh[:mllh_idx] - SIGMA))
u_sig_idx = np.argmin(np.abs(dllh[mllh_idx:] - SIGMA)) + mllh_idx
l_sigma = scan_over[l_sig_idx]
u_sigma = scan_over[u_sig_idx]
else:
l_sigma = f.range[0]
u_sigma = f.range[1]
logging.info('Setting {0} lower sigma bound to '
'{1}'.format(f.name, l_sigma))
f.value = l_sigma
cfx_pipe.update_params(f)
l_out = cfx_pipe.get_outputs()
logging.info('Setting {0} upper sigma bound to '
'{1}'.format(f.name, u_sigma))
f.value = u_sigma
cfx_pipe.update_params(f)
u_out = cfx_pipe.get_outputs()
f.reset()
cfx_pipe.update_params(f)
f_outdir = outdir + '/' + f.name
l_outdir = f_outdir + '/' + 'lower'
u_outdir = f_outdir + '/' + 'upper'
fileio.mkdir(f_outdir)
fileio.mkdir(l_outdir)
fileio.mkdir(u_outdir)
compare(outdir=l_outdir,
ref=MapSet([nom_out]),
ref_label='baseline',
test=MapSet([l_out]),
test_label=r'-sigma',
**args)
compare(outdir=u_outdir,
ref=MapSet([nom_out]),
ref_label='baseline',
test=MapSet([u_out]),
test_label=r'+sigma',
**args)
l_in_mapset = l_outdir + '/' + 'fract_diff__-sigma___baseline.json.bz2'
u_in_mapset = u_outdir + '/' + 'fract_diff__+sigma___baseline.json.bz2'
l_in_map = MapSet.from_json(l_in_mapset).pop() * 100.
u_in_map = MapSet.from_json(u_in_mapset).pop() * 100.
if make_pdf:
outfile = f_outdir + '/systematic_effect.pdf'
else:
outfile = f_outdir + '/systematic_effect.png'
title = r'% effect on ' + r'${0}$'.format(l_in_map.tex) + \
' event counts for {0} parameter'.format(f.name)
sub_titles = (r'(-\sigma - {\rm baseline}) \:/\: {\rm baseline}',
r'(+\sigma - {\rm baseline}) \:/\: {\rm baseline}')
make_plot(
maps=(l_in_map, u_in_map),
outfile=outfile,
logv=False,
center_zero=center_zero,
vlabel=r'({\rm change} - {\rm baseline}) \:/\: {\rm baseline} (%)',
title=title,
sub_titles=sub_titles)
def create_mc_template(toymc_params,
config_file=None,
seed=None,
keep_same_weight=True):
'''
Create MC template out of a pisa pipeline
'''
if seed is not None:
np.random.seed(seed)
Config = parse_pipeline_config(config_file)
# Change binning
Config[('data', 'pi_simple_signal')]['output_specs'] = toymc_params.binning
Config[(
'likelihood',
'pi_generalized_llh_params')]['output_specs'] = toymc_params.binning
# If keep_same_weight is True, turn off the mean adjust and pseudo weight of pi_generalized_llh
if keep_same_weight:
Config[('likelihood',
'pi_generalized_llh_params')]['with_mean_adjust'] = False
Config[('likelihood',
'pi_generalized_llh_params')]['with_pseudo_weight'] = False
else:
Config[('likelihood',
'pi_generalized_llh_params')]['with_mean_adjust'] = True
Config[('likelihood',
'pi_generalized_llh_params')]['with_pseudo_weight'] = True
new_n_events_data = Param(name='n_events_data',
value=toymc_params.n_data,
prior=None,
range=None,
is_fixed=True)
new_sig_frac = Param(name='signal_fraction',
value=toymc_params.signal_fraction,
prior=None,
range=None,
is_fixed=True)
new_stats_factor = Param(name='stats_factor',
value=toymc_params.stats_factor,
prior=None,
range=None,
is_fixed=True)
# These should match the values of the config file, but we override them just in case we need to change these later
new_mu = Param(name='mu',
value=toymc_params.mu,
prior=None,
range=[0, 100],
is_fixed=False)
new_sigma = Param(name='sigma',
value=toymc_params.sigma,
prior=None,
range=None,
is_fixed=True)
Config[('data', 'pi_simple_signal')]['params'].update(p=ParamSet([
new_n_events_data, new_sig_frac, new_stats_factor, new_mu, new_sigma
]))
MCtemplate = DistributionMaker(Config)
return MCtemplate