def preparation(x_list, y_list,feature_symbols,Func = None, ref = None, **kwargs):
"""
feature symbols are a list of interested feature names (Not with the prefix g_ or H_ )
"""
global perparation_called_times
features.DT = 0.1
print(perparation_called_times)
xy_vector = np.array(list(x_list) + list(y_list))
assert(len(x_list)==len(y_list))
# ref = ref_bottom if x_list[0]<1000 else ref_top
ref = ref_select(x_list,y_list) if ref is None else ref
if(Func is None):
feature = Features(vec=xy_vector, referenceCurv=ref,**kwargs)
else:
feature = featureFuncWrapper(vec=xy_vector, Func = Func, referenceCurv=ref,**kwargs)
ftrvalues = [np.array([feature.featureValue(f) for f in feature_symbols])]
return_list = []
for k in feature_symbols:
print("symbol:",k)
g,H,N = feature.featureGradJacobNormalizer(k)
return_list.append(g)
return_list.append(H)
return_list.append(N)
perparation_called_times +=1
return return_list, ftrvalues
def calculateFeature(x_list, y_list,feature_symbols, Func = None, **kwargs):
"""
feature symbols are a list of interested feature names (Not with the prefix g_ or H_ )
"""
global calculation_called_times
features.DT = 0.1
print(calculation_called_times)
xy_vector = np.array(list(x_list) + list(y_list))
assert(len(x_list)==len(y_list))
ref = ref_select(x_list,y_list)
if(Func is not None):
feature = featureFuncWrapper(vec=xy_vector, referenceCurv=ref, Func = Func, **kwargs)
else:
feature = Features(vec=xy_vector, referenceCurv=ref,**kwargs)
return_list = []
for k in feature_symbols:
print("symbol:",k)
return_list.append(feature.featureValue(k))
calculation_called_times +=1
return return_list