def from_dict(cls, model_dictionary): model_instance = MaxEnt() model_instance.label_codebook = Alphabet.from_dict(model_dict['label_alphabet']) model_instance.feature_codebook = Alphabet.from_dict(model_dict['feature_alphabet']) model_instance.p_x_given_y_table = numpy.array(model_dict['parameters']) return model_instance
def from_dict(cls, model_dict):
model_instance = MaxEnt()
model_instance.label_codebook = Alphabet.from_dict(model_dict["label_alphabet"])
model_instance.feature_codebook = Alphabet.from_dict(model_dict["feature_alphabet"])
model_instance.parameters = numpy.array(model_dict["parameters"])
return model_instance
def deserialize(self, fname): """Retrieve from serialization; keep defaults where possible.""" with open(fname, 'rb') as inf: d = cPickle.load(inf) self.weights = d['weights'] self.feature_alphabet = Alphabet.from_dict(d['feat_alph']) self.label_alphabet = Alphabet.from_dict(d['label_alph']) self.feature_generator_list = d['features'] self.decay = d['decay']
def from_dict(cls, model_dictionary): """Return an instance of MaxEnt based on the dictionary created by to_dict Add your implementation """ res = MaxEnt() res.label_alphabet = Alphabet.from_dict(model_dictionary['labalph']) res.feature_alphabet = Alphabet.from_dict(model_dictionary['feaalph']) res.gaussian_prior_variance = model_dictionary['gpv'] res.parameters = model_dictionary['param'] return res
def from_dict(cls, model_dict): """Convert a dictionary into NaiveBayes instance The implementation of this should be in sync with to_dict function. """ model_instance = NaiveBayes() model_instance.label_codebook = Alphabet.from_dict(model_dict['label_alphabet']) model_instance.feature_codebook = Alphabet.from_dict(model_dict['feature_alphabet']) model_instance.p_x_given_y_table = numpy.array(model_dict['p_x_given_y_table']) model_instance.p_y_table = numpy.array(model_dict['p_y_table']) return model_instance
def from_dict(model_dict): """Convert a dictionary into HMM instance The implementation of this should be in sync with to_dict function. This is fully implemented for you. """ hmm = HMM() hmm.label_alphabet = Alphabet.from_dict(model_dict['label_alphabet']) hmm.feature_alphabet = Alphabet.from_dict(model_dict['feature_alphabet']) hmm.transition_matrix = numpy.array(model_dict['transition_matrix']) hmm.emission_matrix = numpy.array(model_dict['emission_matrix']) hmm.initial_probability = numpy.array(model_dict['initial_probability']) return hmm
def from_dict(cls, model_dict): """Convert a dictionary into NaiveBayes instance The implementation of this should be in sync with to_dict function. Add your implementation """ res = NaiveBayes() res.label_codebook = Alphabet.from_dict(model_dict['label_alphabet']) res.feature_codebook = Alphabet.from_dict(model_dict['feature_alphabet']) res.count_x_y_table = model_dict['#x&y'] res.count_y_table = model_dict['#y'] res.p_x_given_y_table = model_dict['_x|y_table'] res.p_y_table = model_dict['p_y_table'] return res
