Dataset.py

import os
import numpy as np
from sklearn.model_selection import train_test_split


class Dataset:
    def loadDataset(self):  # load the gray data that was read before and labeled
        return np.load("Data/DatasetGray.npy", allow_pickle=True)  # load dataset

    def saveDataset(
        self, path, data
    ):  # save the data of images with labels to be used later
        np.save(path, data)

    def loadTrainData(self):
        if os.path.exists("Data/x_train.npy"):
            images = np.load("Data/x_train.npy", allow_pickle=True)
        if os.path.exists("Data/y_train.npy"):
            labels = np.load("Data/y_train.npy", allow_pickle=True)
        return images, labels

    def loadTestData(self):
        if os.path.exists("Data/x_test.npy"):
            images = np.load("Data/x_test.npy", allow_pickle=True)
        if os.path.exists("Data/y_test.npy"):
            labels = np.load("Data/y_test.npy", allow_pickle=True)
        return images, labels

    def CreateValidData(self):
        x_train, y_train = self.loadTrainData()
        x_train, x_valid, y_train, y_valid = train_test_split(
            x_train, y_train, test_size=0.2
        )
        return x_train, x_valid, y_train, y_valid