Trainer.py

from CAModel import CAModel
from DataGenerator import DataGenerator

from sklearn.metrics import accuracy_score
from sklearn.metrics import multilabel_confusion_matrix, classification_report, confusion_matrix
import numpy as np

class Trainer():
    """
    A class used to create our Keras model
    """

    def __init__(self,gen_train, gen_eval,gen_test,label_test) -> None:
        self.gen_train = gen_train
        self.gen_eval = gen_eval

        self.gen_test = gen_test
        self.label_test = label_test

        self.caModel = CAModel()
        return
    
    def train(self):
        """Returns a trained model"""
        model = self.caModel.getMode()
        #there are better techniques to train this 

        #we need to figure out how to turn the nucleotides into numbers for this to work
        model.fit(self.gen_train, validation_data=self.gen_eval, epochs=20, steps_per_epoch=1000)

        model.save('models/large_no_conv1d')
        validation_data_results = model.predict(self.gen_test)

        validation_data_results.round()

        y_test_arg=np.argmax(self.label_test,axis=1)
        Y_pred = np.argmax(model.predict(self.gen_test),axis=1)

        print("predict sample: ", validation_data_results[0])
        print("predict sample 2: ", validation_data_results[1])
        print("predict sample 3: ", validation_data_results[2])
        print("predict sample 4: ", validation_data_results[3])


        print('Confusion Matrix')
        print(confusion_matrix(y_test_arg, Y_pred))
        
        print('Accuracy')
        print(accuracy_score(y_test_arg,Y_pred))

        return model