Untitled

class MY_SparsePCA():
    def __init__(self, n_components, alpha, max_iter=1000, tol=1e-6):
        self.alpha = alpha
        self.n_components = n_components
        self.max_iter = max_iter
        self.tol = tol
        self.components_ = None
        self.error_ = None
        self.S = None
        self.Vt = None

    @staticmethod
    def soft_threshold(x, alpha):
        # GRADED CODE: Iterative Thresholding
        ### START CODE HERE ###
        # Soft threshold function used for SPCA.
        rt = None
        ### END CODE HERE ###

        return rt

    def fit_transform(self, x):

        # GRADED CODE: SPARSE PCA FITTINT FUNCTION
        ### START CODE HERE ###
        # Initialize components (5%)
        _, S, Vt = np.linalg.svd(x, full_matrices=False)
        self.components_ = None
        self.S = S
        self.Vt = Vt

        for iteration in tqdm(range(self.max_iter)):

            # Compute the transform in this iteration
            x_projected = None

            # Iterative Thresholding (update the components) (5%):
            self.components_ = None

            # Normalize components (5%)
            norm = None
            norm[norm < 1e-10] = 1.0
            self.components_ /= None

            # SSD Check
            ssd = None
            if ssd < self.tol:
                break
        # Transform the data (5%)
        x_transformed = None

        ### END CODE HERE###

        return x_transformed

    def fit(self, x):
        self.fit_transform(x)
        return self

    def transform(self, x):
        return x@self.components_