194b2c59307ea8f46b55840d4fd92bf85173f168,textClassification/wrapper.py,Classifier,predict,#Classifier#Any#Any#,88

Before Change

        if self.model_config.fold_number is 1:
            if self.model is not None:
                //classifier = Classifier(self.model, preprocessor=self.p)
                x_t = self.p.to_sequence(texts, maxlen=300)
                result = predict(self.model, x_t)
            else:
                raise (OSError("Could not find a model."))
        else:
            if self.models is not None:
                x_t = self.p.to_sequence(texts, maxlen=300)
                result = predict_folds(self.models, x_t)
            else:
                raise (OSError("Could not find nfolds models."))
        if output_format is "json":

After Change

        self.models = train_folds(x_train, y_train, self.model_config, self.training_config, self.embeddings)

    // classification
    def predict(self, texts, output_format="json"):
        if self.model_config.fold_number is 1:
            if self.model is not None:
                //classifier = Classifier(self.model, preprocessor=self.p)
                //x_t = self.p.to_sequence(texts, maxlen=300)
                predict_generator = DataGenerator(texts, None, batch_size=self.model_config.batch_size, 
                    maxlen=self.model_config.maxlen, list_classes=self.model_config.list_classes, 
                    embed_size=self.model_config.word_embedding_size, embeddings=self.embeddings, shuffle=False)

                //x_t = self.p.to_vector(texts, self.embeddings, maxlen=self.model_config.maxlen, embed_size=self.model_config.word_embedding_size)
                result = predict(self.model, predict_generator)
            else:
                raise (OSError("Could not find a model."))
        else:
            if self.models is not None:
                //x_t = self.p.to_sequence(texts, maxlen=300)
                //x_t = self.p.to_vector(texts, self.embeddings, maxlen=self.model_config.maxlen, embed_size=self.model_config.word_embedding_size)
                predict_generator = DataGenerator(texts, None, batch_size=self.model_config.batch_size, 
                    maxlen=self.model_config.maxlen, list_classes=self.model_config.list_classes, 
                    embed_size=self.model_config.word_embedding_size, embeddings=self.embeddings, shuffle=False)

                result = predict_folds(self.models, predict_generator)
            else:
                raise (OSError("Could not find nfolds models."))

Instances