def calcweight(self, causevardata, affectedvardata, *_):
""""Calculates the transfer entropy between two vectors containing
timer series data.
"""
# Calculate transfer entropy as the difference
# between the forward and backwards entropy
# Pass special estimator specific parameters in here
transent_fwd, auxdata_fwd = transentropy.calc_infodynamics_te(
self.infodynamicsloc,
self.estimator,
affectedvardata.T,
causevardata.T,
**self.parameters
)
transent_bwd, auxdata_bwd = transentropy.calc_infodynamics_te(
self.infodynamicsloc,
self.estimator,
causevardata.T,
affectedvardata.T,
**self.parameters
)
transent_directional = transent_fwd - transent_bwd
transent_absolute = transent_fwd
return [transent_directional, transent_absolute], [auxdata_fwd, auxdata_bwd]
def calcweight(self, causevardata, affectedvardata, weightcalcdata,
causevarindex, affectedvarindex):
""""Calculates the transfer entropy between two vectors containing
timer series data.
"""
# Calculate transfer entropy as the difference
# between the forward and backwards entropy
# Pass special estimator specific parameters in here
transent_fwd, auxdata_fwd = \
transentropy.calc_infodynamics_te(
self.infodynamicsloc, self.estimator,
affectedvardata.T, causevardata.T, **self.parameters)
transent_bwd, auxdata_bwd = \
transentropy.calc_infodynamics_te(
self.infodynamicsloc, self.estimator,
causevardata.T, affectedvardata.T, **self.parameters)
transent_directional = transent_fwd - transent_bwd
transent_absolute = transent_fwd
return [transent_directional, transent_absolute], \
[auxdata_fwd, auxdata_bwd]
def tecalc_wrapper(causevardata, affectedvardata):
te_fwd = calc_infodynamics_te(infodynamics_loc, estimator, causevardata.T,
affectedvardata.T, **add_parameters)
te_bwd = calc_infodynamics_te(infodynamics_loc, estimator,
affectedvardata.T, causevardata.T,
**add_parameters)
return [te_fwd, te_bwd]
def tecalc_wrapper(causevardata, affectedvardata):
te_fwd = calc_infodynamics_te(
infodynamics_loc, estimator, causevardata.T, affectedvardata.T, **add_parameters
)
te_bwd = calc_infodynamics_te(
infodynamics_loc, estimator, affectedvardata.T, causevardata.T, **add_parameters
)
return [te_fwd, te_bwd]
def calc_surr_te(self, affected_data, causal_data, num):
"""Calculates surrogate transfer entropy values for significance
threshold purposes.
Two methods for generating surrogate data is available:
iAAFT (Schreiber 2000a) or random_shuffle in time.
Returns list of surrogate transfer entropy values of length num.
"""
# The causal (or source) data is replaced by surrogate data,
# while the affected (or destination) data remains unchanged.
# Get the causal data in the correct format
# for surrogate generation
original_causal = np.zeros((1, len(causal_data)))
original_causal[0, :] = causal_data
if self.te_surr_method == 'iAAFT':
surr_tsdata = \
[data_processing.gen_iaaft_surrogates(
original_causal, 10)
for n in range(num)]
elif self.te_surr_method == 'random_shuffle':
surr_tsdata = \
[data_processing.shuffle_data(causal_data) for n in range(num)]
surr_te_fwd = []
surr_te_bwd = []
for n in range(num):
surr_te_fwd.append(
transentropy.calc_infodynamics_te(self.infodynamicsloc,
self.estimator,
affected_data,
surr_tsdata[n][0, :],
**self.parameters)[0])
surr_te_bwd.append(
transentropy.calc_infodynamics_te(self.infodynamicsloc,
self.estimator,
surr_tsdata[n][0, :],
affected_data,
**self.parameters)[0])
surr_te_directional = \
[surr_te_fwd[n] - surr_te_bwd[n] for n in range(num)]
surr_te_absolute = [surr_te_fwd[n] for n in range(num)]
return surr_te_directional, surr_te_absolute